Monday

8:15 Continental Breakfast
Channel Capacity
Chair: Vinod Sharma, IISc
Distributed Computing
Chair: Emina Soljanin, Rutgers
Wireless Networks
Chair: Andrea Goldsmith, Stanford University
Signal Processing
Chair: Junting Chen, USC
Private Information
Chair: Prakash Narayan, University of Maryland
Deep Learning Theory
Chair: Sean Meyn, University of Florida
9:00 Capacity upper bounds for deletion-type channels
Mahdi Cheraghchi*, Imperial College London
 
A combinatorial design of coded distributed computing
Nicholas Woolsey, University Of Utah; Rong-Rong Chen, University Of Utah; Mingyue Ji*, University Of Utah
 
Maximum area spectral efficiency for extremely dense cellular networks
Ahmad Alammouri, UT Austin; Jeffrey Andrews*, UT Austin; Francois Baccelli, UT Austin
 
On feature learning with state space models and pulse domain signal analysis
Hans-Andrea Loeliger*, ETH; Sarah Neff, ETH; Nour Zalmai, ETH
 
A capacity-achieving PIR protocol for distributed storage using an arbitrary linear code
Hsuan-Yin Lin, Simula@UiB; Siddhartha Kumar, Simula@UiB; Eirik Rosnes*, Simula@UiB; Alexandre Graell i Amat, Chalmers University Of Technology Sweden
 
Deep neural network architectures for modulation classification
Xiaoyu Liu, Purdue; Diyu Yang, Purdue; Aly El Gamal*, Purdue
 
9:20 Log ratio of entropy powers
Jerry Gibson*, UC Santa Barbara
 
Communication-aware scheduling of serial tasks for dispersed computing
Chien-Sheng Yang, USC; Ramtin Pedarsani*, UC Santa Barbara; Salman Avestimehr, USC
 
On random access with energy harvesting nodes
Tolga M. Duman*, Bilkent University, Turkey; Umut Demirhan, Arizona State University
 
A parametric approach for graph system identification
Anna Scaglione*, Arizona State; Hoi To Wai, Arizona State; Amir Leshem, Bar-Ilan University; Asuman Ozgladar, MIT
 
Key agreement through broadcast channel measurements
Onur Günlü, TU Munich; Rafael Schaefer*, Technische Universität Berlin
 
Fit without fear: A perspective on deep and shallow learning
Mikhail Belkin*, Ohio State; Siyuan Ma, Ohio State; Soumik Mandal, Ohio State
 
9:40 Graph-based auxiliary random variable and its use for feedback capacity
Haim Permuter*, Ben Gurion University; Oron Sabag
 
Entangled polynomial codes for high-dimensional distributed matrix multiplication
Qian Yu, USC; Mohammad Maddah-Ali, AT&T Bell Labs; Salman Avestimehr*, USC
 
Beamforming with multiple one-bit wireless transceivers
Kang Gao, University Of Notre Dame; J Nicholas Laneman, University Of Notre Dame; Bertrand Hochwald*, University Of Notre Dame
 
5G ultra-reliable and low-latency systems design
Chih-Ping Li, Qualcomm; Tingfang Ji, Qualcomm; Jing Jiang*, Qualcomm; Wei Yang, Qualcomm; John Smee, Qualcomm; Wanshi Chen, Qualcomm
 
Private information retrieval with side information
Swanand Kadhe, University of California, Berkeley; Brenden Garcia, Texas A&M University; Anoosheh Heidarzadeh, Texas A&M University; Salim El Rouayheb, Texas A&M University; Alex Sprintson*, Texas A&M University
 
GANs for compressed sensing, classification and defense
Alex Dimakis*, UT Austin; Ashish Bora, University Of Texas; Eric Price, University Of Texas; Ajil Jalal, University Of Texas
 
10:00 On asymptotic analysis of energy-distortion tradeoff for low-delay transmission over Gaussian channels
Ceren Sevinc, UC Riverside; Ertem Tuncel*, UC Riverside
 
Edge-Assisted Autonomous Systems
Marco Levorato*, UC Irvine
 
Uplink oblivious cloud radio access networks: An information theoretic overview
I. Estella Aguerri, Huawei; Abdellatif Zaidi, University Of Paris-Est; Giuseppe Caire, TU Of Berlin; Shlomo Shamai*, Technion
 
Neural network pruning: deep learning and compressed sensing meet
Alireza Aghasi*, Georgia State Business School
 
Private information retrieval under constraints
Karim Banawan, University Of Maryland; Sennur Ulukus*, University Of Maryland
 

 
10:20 Break
Channel Coding 1
Chair: Michael Langberg, SUNY Buffalo
Coding Theory 1
Chair: Daniel Costello, University of Notre Dame
Modulation
Chair: Joseph J. Boutros, Texas A&M at Qatar
Privacy and Security
Chair: Alex Sprintson, Texas A&M
Deep and Large Scale Learning
Chair: Alex Dimakis, UT Austin
Estimation and Prediction
Chair: Aria Nosratinia, UT Dallas
10:40 The channel capacity from particles to waves
Richard Blahut*, UIUC; George Papen, UC San Diego
 
On skip sliding window codes
Ting-Yi Wu, UIUC; Anshoo Tandon, National University Of Singapore; Lav Varshney, UIUC; Mehul Motani*, National University Of Singapore
 
On OTFS modulation for high-doppler fading channels
K. R. Murali, Indian Institute Of Science; A. Chockalingam*, Indian Institute Of Science
 
Social network de-anonymization based on group memberships: An information theoretic approach
Farhad Shirani, NYU; Siddharth Garg, NYU; Elza Erkip*, NYU
 
Accelerating stochastic gradient descent
Rahul Kidambi, University Of Washington; Praneeth Netrapalli, Microsoft; Prateek Jain*, Microsoft; Sham Kakade, University Of Washington
 
Absorbing Markov chains for public security
Ninoslav Marina*, University of Information Science and Techniology; Magdalena Puncheva, HES-SO
 
11:00 On the zero-error capacity of channels with noisy feedback
Natasha Devroye*, University Of Illinois Chicago; Meysam Asadi, University Of Illinois Chicago
 
Error control of line codes generated by finite Coxeter groups
Ezio Biglieri*, Pompeu Fabra University; Emanuele Viterbo, Monash University; Michele Elia, Politecnico Di Torino
 
Some results on coded optical CDMA
Dariush Divsalar*, JPL/Caltech; Matthew Thill, JPL/Caltech; Sam Dolinar, JPL/Caltech; Sudarsan Ranganathan, UCLA; Richard Wesel, UCLA
 
Secure private information retrieval from distributed databases
Jungwoo Lee*, Seoul National University; Heecheol Yang, Seoul National University
 
Learning deep models: Critical points and local openness
Maher Nouiehed, USC; Meisam Razaviyayn*, USC
 
Bilinear estimation using EM vector AMP
Phil Schniter*, Ohio State; Subrata Sarkar, Ohio State
 
11:20 Limited feedback design for non-orthogonal multiple access
Hamid Jafarkhani*, UC Irvine
 
Recent results on the GM-MDS conjecture or: How to construct MDS codes with support constraints over small fields
Anoosheh Heidarzadeh*, Texas A&M; Alex Sprintson, Texas A&M
 
Enumeration of paraunitary-based complementary QAM sequences
Predrag Spasojevic*, Rutgers; Srdjan Budishin, RT-RK, Serbia
 
Privacy issues arising in spectrum sharing between radar and communications systems
Anastasios Dimas, Rutgers; Matthew Clark, Qualcomm; Konstantinos Psounis, USC; Athina Petropulu*, Rutgers
 
Fast distributed learning in the presence of slow and stale updates
Sanghamitra Dutta, CMU; Gauri Joshi*, CMU; Soumyadip Ghosh, IBM T.J Watson; Parijat Dube, IBM T.J Watson; Priya Nagpurkar, IBM T.J Watson
 

 
11:40 Coded positioning systems
Lele Wang*, Stanford; Ofer Shayevitz, Tel Aviv University; Tsachy Weissman, Stanford
 
Immink coding and state splitting method
Shan Lu, Gifu University; Hiroshi Kamabe*, Gifu University
 
Efficient detection of OTFS modulation
Patchava Raviteja, Monash University; Khoa Tran Phan, Monash University; Yi Hong, Monash University; Emanuele Viterbo*, Monash University
 
Price of anonymity in heterogeneous distributed detection
Wei-Ning Chen, National Taiwan University; I-Hsiang Wang*, National Taiwan University
 
Robust distributed learning in the face of adversity
Dimitris Papailiopoulos*, University Of Wisconsin
 

 
12:00 Lunch
1:30 Plenary: Sub-Nyquist Sampling without Sparsity, Yonina Eldar, The Technion
2:15 Plenary: Millimeter Wave MIMO Signal Processing, Robert Heath, University of Texas at Austin
3:00 Break
Source Coding
Chair: Jean-Francois Chamberland, Texas A&M
Coding For Storage 1
Chair: Eitan Bachmat, BGU
MIMO
Chair: Nicolo Michelusi, Purdue
Compressed Sensing
Chair: Deanna Needell, UCLA
Learning in High Dimensions
Chair: Varun Jog, UW Madison

3:15 Remote source coding under Gaussian noise : Dueling roles of power and entropy-power
Krishnan Eswaran, UC Berkeley and Google; Michael Gastpar*, EPFL
 
Block-diagonal and LT codes for distributed computing with straggling servers
Albin Severinson, Simula@UiB; Alexandre Graell i Amat*, Chalmers University of Technology; Eirik Rosnes, Simula@UiB
 
Practical integer-forcing MIMO transceivers for wireless communications
Sung Ho Chae, Samsung; Kwang Taik Kim*, Purdue; Seok-Ki Ahn, Samsung; Jeongho Park, Samsung; Young-Han Kim, UC San Diego
 
On compressed sensing based schemes for massive multiple access
Vamsi Amalladinne, Texas A&M University; Avinash Vem, Nvidia; Dileep Soma, Texas A&M University; Krishna Narayanan*, Texas A&M University; Jean-Francois Chamberland, Texas A&M University
 
A large dimensional analysis of regularized discriminant analysis classifiers
Khalil Elkhalil, KAUST; Abla Kammoun, KAUST; Romain Couillet, CNRS-Centrale Supelec-Universite Paris-Sud; Tareq Al-Naffouri*, KAUST; Mohamed-Slim Alouini, KAUST
 

 
3:35 Ozarow-type outer bounds for sources and channels
Yuval Kochman, HUJI; Or Ordentlich*, HUJI; Yury Polyanskiy, MIT
 
Optimal repair of Reed-Solomon codes
Min Ye, Princeton; Itzhak Tamo, Tel Aviv University; Alexander Barg*, University Of Maryland
 
Error control sounding strategies for millimeter wave beam alignment
Vinayak Suresh, Purdue; David Love*, Purdue
 
Joint factorization and decomposition of compressed tensors for low-rank tensor recovery and some applications
Mohsen Joneidi, University of Central Florida; Nazanin Rahnavard*, University of Central Florida
 
Learning structured dictionaries for multidimensional data
Mohsen Ghassemi, Rutgers; Zahra Shakeri, Rutgers; Waheed Bajwa, Rutgers; Anand Sarwate*
 

 
3:55 An LP upper bound for rate distortion with variable side information
Sinem Unal, Kencast; Aaron Wagner*, Cornell
 
Coding over sets for DNA storage
Andreas Lenz, TU Munich; Paul Siegel, UC San Diego; Antonia Wachter-Zeh, TU Munich; Eitan Yaakobi*, Technion
 
Adaptive channel estimators with hybrid beamforming for single-carrier massive MIMO
Sadjad Sedighi, UC Irvine; Gokhan M. Guvensen, METU; Ender Ayanoglu*, UC Irvine
 
Fast binary embeddings, and quantized compressed sensing with structured matrices
Thang Huynh*, UC San Diego; Rayan Saab, UC San Diego
 
An information-theoretic view of learning in high dimensions: Universal features, maximal correlations, bottlenecks, and common information
Shao-Lun Huang, TBSI; Anuran Makur, MIT; Gregory Wornell*, MIT; Lizhong Zheng, MIT
 

 
4:15 Break
Caching
Chair: Urs Niesen, Qualcomm
LDPC Codes
Chair: David Mitchell, New Mexico State
Energy Harvesting
Chair: Tolga M. Duman, Bilkent University
Quantum and Beyond
Chair: Haim Permuter, BGU
Hypothesis Testing
Chair: Himanshu Tyagi, IISc

4:30 Adding one transmitter can double the effective gains of coded caching
Petros Elia*, EURECOM; Eleftherios Lampiris
 
Improved trellis-based algorithm for locating and breaking cycles in bipartite graphs with applications to LDPC codes
Juane Li, Appotech, USA; Shu Lin, UC Davis; Khaled Abdel-Ghaffar*, UC Davis
 
Age-minimal online policies for energy harvesting sensors with incremental battery recharges
Ahmed Arafa*, Princeton; Jing Yang, Penn State; Sennur Ulukus, University Of Maryland College Park; H. Vincent Poor, Princeton
 
Modified RM code-based signature schemes
Wijik Lee, Seoul National University; Jong-Seon No*, Seoul National University; Young-Sik Kim, Chosun University; Yongwoo Lee, Seoul National University
 
On the asymptotic blocklength-dimension tradeoff of composite hypothesis testing
Michael Bell, Hebrew University; Yuval Kochman*, Hebrew University
 

 
4:50 Recent advances on combination networks with end-user caches
Daniela Tuninetti - UIC*, University Of Illinois Chicago; Pablo Piantanida, Supelec; Mingyue Ji, University Of Utah; Kai Wan, Supelec
 
Reed-Solomon-based quasi-cyclic LDPC codes: Designs, cycle structure, and erasure correction
Xin Xiao, University Of Arizona; William Ryan*, Zeta Associates; Bane Vasic, University Of Arizona; Shu Lin, UC Davis; Khaled Abdel-Ghaffar, UC Davis
 
Intermittent Kalman filtering with an energy harvesting sensor
Omur Ozel*, CMU; Venkat Anantharam, UC Berkeley
 
Opening the quantum box: A probabilistic perspective on quantum annealers
Andrey Lokhov, Los Alamos National Laboratory; Carleton Coffrin, Los Alamos National Laboratory; Sidhant Misra, Los Alamos National Laboratory; Marc Vuffray*, Los Alamos National Laboratory
 
Distributed hypothesis testing over multi-access channels
Sadaf Salehkalaibar, Tehran University; Michele Wigger*, Telecom ParisTech
 

 
5:10 On coded caching with heterogeneous distortion requirements
Abdelrahman Ibrahim, Penn State; Ahmed Zewail, Penn State; Aylin Yener*, Penn State
 
Learning to decode LDPC codes with finite-alphabet message passing
Xin Xiao, University Of Arizona; Bane Vasic*, University Of Arizona; Shu Lin, UC Davis
 
Minimizing age of information for an energy harvesting source with updating error
Songtao Feng, Penn State; Jing Yang*, Penn State
 

 
Controlled sensing for hypothesis testing under correlated actions
Javad Heydari, RPI; Ali Tajer*, RPI
 

 

Tuesday

8:15 Continental Breakfast
Channel Capacity 1
Chair: Mahdi Cheraghchi, Imperial College
Source Coding 2
Chair: Or Ordentlich, HU Jerusalem
5G
Chair: Jeffrey Andrews, UT Austin
Communication
Chair: Predrag Spasojevic, Rutgers
Fairness and Learning
Chair: Anand Sarwate, Rutgers
Learning Theory
Chair: Mesrob Ohannessian, TTIC
9:00 On the communication consequences of a rate-0 edge
Michelle Effros*, Caltech
 
Joint source-channel coding with concatenated spatially coupled LDPC codes
Ahmad Golmohammadi, New Mexico State; David Mitchell*, New Mexico State
 
Taking 5G to the skies: designing millimeter wave systems for A2A channels
Nuria Gonzalez Prelcic, University of Vigo; Javier Rodriguez Fernandez, University of Vigo; Robert Heath*, University Of Texas
 
Interference management for full-duplex cellular networks
Sang-Woon Jeon*, Hanyang University; Minho Yang, Yonsei University; Dong Ku Kim, Yonsei University
 
Your dreams may come true with MTP2
Caroline Uhler*, MIT; Piotr Zwiernik, Pompeu Fabra University; Steffen Lauritzen, University Of Copenhagen
 
Interactive algorithms for graphical model selection
Gautam Dasarathy*, Rice; Aarti Singh, CMU; Maria-Florina Balcan, CMU; Richard Baraniuk, Rice
 
9:20 Feedback capacity of stationary Gaussian channels further examined
Tao Liu, University Of Hong Kong; Guangyue Han*, University Of Hong Kong
 
Batch learning with log-loss
Meir Feder*, Tel Aviv University; Yaniv Fogel, Tel Aviv University
 
Toward a user-centric analysis of 5G networks
Martin Haenggi*, University Of Notre Dame
 
A random access scheme to balance energy efficiency and accuracy in monitoring applications
Michele Zorzi*, University Of Padova; Chiara Pielli, University Of Padova; Daniel Zucchetto, University Of Padova; Andrea Zanella, University Of Padova
 
Calibration for the (computationally-identifiable) masses
Úrsula Hébert-Johnson, Stanford; Michael P. Kim, Stanford; Omer Reingold*, Stanford; Guy N. Rothblum, Weizmann
 
Learning for policy improvement
Geoffrey Gordon*, Microsoft Research and CMU
 
9:40 Bounds on capacity of a non-linear fiber optic channel with memory
Konchady Gautam Shenoy, Indian Institute Of Science Bangalore India; Vinod Sharma*, Indian Institute Of Science Bangalore India
 
Information aging through queues: A mutual information perspective
Yin Sun*, Auburn University
 
5G opportunities and challenges.
Reinaldo Valenzuela*, Nokia, Bell Labs; Rodolfo Feick, UTFSM; Dmitry Chizhik, Nokia, Bell Labs; Jinfeng Du, Nokia, Bell Labs; Mauricio Rodriguez, UCV
 
Efficient multidimensional wideband parameter estimation for OFDM based joint radar and communication systems
Martin Haardt*, TU Ilmenau; Ivan Podkurkov, TU Ilmenau; Jianshu Zhang, TU Ilmenau; Adel Nadeev, KNRTU Kazan
 
On optimal generalizability in parametric learning
Ahmad Beirami*, MIT; Meisam Razaviyayn, USC; Shahin Shahrampour, Harvard; Vahid Tarokh, Duke
 
Zap stochastic approximation for fastest reinforcement learning
Adithya Devraj, University Of Florida; Ana Busic, Inria Paris; Sean Meyn*, University Of Florida
 
10:00 Participation incentives on a wireless random access erasure collision channel
Vasilis Gkatzelis, Drexel University; Steven Weber*, Drexel University
 
Empirical Kolmogorov complexity
Ari Trachtenberg*, BU
 
Sparse vector coding for ultra short packet transmission
Hyoungju Ji, Seoul National University; Byonghyo Shim*, Seoul National University; Sangtae Kim, Seoul National University
 
Low-latency FBMC receivers
Jeremy Nadal, TELECOM Bretagne; Francois Leduc-Primeau*, Ecole de Technologie Superieure; Charbel Abdel Nour, TELECOM Bretagne; Amer Baghdadi, TELECOM Bretagne
 
Fairness in supervised learning: An information theoretic approach
AmirEmad Ghassami, UIUC; Sajad Khodadadian, UIUC; Negar Kiyavash*, UIUC
 
Learning the topology of a network via graph filter identification
Antonio Marques*, King Juan Carlos University; Santiago Segarra, MIT; Gonzalo Mateos, University Of Rochester; Rasoul Shafipour, University Of Rochester
 
10:20 Break
Network Information Theory
Channel Capacity 2
Chair: Richard Blahut, UIUC
MM Wave
Chair: David Love, Purdue
Wireless Communication
Chair: Larry Milstein, UC San Diego
Information Theoretic Learning

10:40 Symmetric generalized Gaussian multiterminal source coding
Jun Chen*, McMaster University; Li Xie ; Yameng Chang, McMaster University; Jia Wang, Shanghai Jiao Tong University; Yizhong Wang, Tianjin University of Science and Technology
 
Coded random access for communication of short messages
Mohammadreza Ebrahimi ; Farshad Lahouti*, Caltech; Victoria Kostina, Caltech
 
The information and wave-theoretic limits of analog beamforming
Amine Mezghani*, UT Austin; Robert W. Heath, UT Austin
 
Can machine learning trump theory in communication system design?
Andrea Goldsmith*, Stanford; Nariman Farsad
 
On exponentially concave functions and their impact in information theory
Gholamreza Alirezaei, RWTH Aachen University; Rudolf Mathar*, RWTH Aachen University
 

 
11:00 Applications and distance measures of subspace codes
Tuvi Etzion*, Technion
 
Computing AWGN channel thresholds of spatially coupled turbo-like codes
Muhammad Umar Farooq, Lund University; Saeedeh Moloudi, Lund University; Michael Lentmaier*, Lund University
 
Using source and channel coding for beam discovery in millimeter wave
C. Emre Koksal*, Ohio State; Yahia Shabara, Ohio State; Eylem Ekici, Ohio State
 
BATS: Network coding in action (an update)
Shenghao Yang, The Chinese University of Hong Kong, Shenzhen; Raymond Yeung*, The Chinese University of Hong Kong
 
Single-community detection with side information: information theoretic limits
Aria Nosratinia*, University of Texas at Dallas
 

 
11:20 Finite blocklength converses in point-to-point and network information theory: A convex analytic perspective
Ankur A. Kulkarni*, Indian Institute of Technology Bombay; Sharu Jose, Indian Institute of Technology Bombay
 
Gaussian Interference channel: some results and conjectures
Chandra Nair*, CUHK; David Ng, CUHK
 
Multi-user Beam-Alignment for Millimeter-Wave Networks
Rana A. Hassan, Purdue; Nicolo Michelusi*, Purdue
 
Iterative frequency-domain receiver processing in doubly selective fading channels
Christian Schlegel*, Dalhousie University; Dmitry Truhachev, Dalhousie University; Amir Vaezi, Dalhousie University; Murwan Bashir, Dalhousie University
 
On data-dependent random features for improved generalization in supervised learning
Shahin Shahrampour*, Harvard; Ahmad Beirami, Harvard; Vahid Tarokh, Harvard
 

 
11:40 Dependence balance in multiple access channels with correlated sources
Amos Lapidoth, ETH; Shirin Saeedi Bidokhti*, UPenn; Michele Wigger, Telecom ParisTech
 
Coupling data transmission over MAC: Asymptotics of the achievable rates
Dmitry Trukhachev*, Dalhousie Univ.
 
One-bit precoding and constellation range design for massive MIMO with QAM signaling
Foad Sohrabi, University Of Toronto; Ya-Feng Liu, Chinese Academy of Sciences; Wei Yu*, University Of Toronto
 
Representations of large scale wireless networks for policy optimization
Urbashi Mitra*, USC; Libin Liu, USC; Arpan Chattopadhyay, USC
 
Spectral estimation for non-stationary processes with applications to stationarity test
Yu Xiang*, Harvard; Jie Ding, Harvard; Vahid Tarokh, Harvard
 

 
12:00 Lunch
1:30 Plenary: Sample and Computationally Efficient Active Learning, Nina Balcan, Carnegie Mellon University
2:15 Plenary: Active Machine Learning: From Theory to Practice, Robert Nowak, University of Wisconsin
3:00 Break
Channel Decoding and Capacity
Chair: Natasha Devroye, UIC
Distributed Storage
Chair: Eitan Yaakobi, Technion
Coding Theory 2
Chair: Mehul Motani, NUS
Sequences and Functions
Chair: Ulrich Tamm, FH Bielefeld
Communication Channels
Chair: Shlomo Shamai, Technion
Information and Control
Chair: Vivek Borkar, IIT Bombay
3:15 List decoding as a first step towards unique decoding
Michael Langberg*, SUNY Buffalo
 
Optics, optimization (and flash drive coding)
Eitan Bachmat*, Ben Gurion University
 
Improving the locality of generalized integrated interleaved codes
Xinmiao Zhang*, Ohio State
 
Optimal interactive coding for insertions, deletions, and substitutions
Alexander Sherstov*, UC Los Angeles; Pei Wu, UC Los Angeles
 
Relay for data: An underwater race
Yashar Aval, Northeastern University; Milica Stojanovic, Northeastern University; Sarah Kate Wilson*, Santa Clara University
 
Exact minimum number of bits to stabilize a linear system
Victoria Kostina*, Caltech; Yuval Peres, Microsoft; Gireeja Ranade, Microsoft; Mark Sellke, Cambridge
 
3:35 Efficient binomial channel capacity computation with an application to molecular communication
Richard Wesel*, UC Los Angeles; Emily Wesel, Harvard-Westlake; Lieven Vandenberghe, UC Los Angeles; Christos Komninakis, Qualcomm; Muriel Medard, MIT
 
Erasure correction of scalar codes in the presence of stragglers
Netanel Raviv*, Caltech; Yuval Cassuto, Technion; Rami Cohen, Technion; Moshe Schwartz, Ben Gurion University
 
On decoding error propagation for braided convolutional codes
Min Zhu, Xidian University; David Mitchell, New Mexico State; Michael Lentmaier, Lund University; Daniel Costello*, University Of Notre Dame
 
Recent development on perfect polyphase sequences and optimal familes
Min Kyu Song, Yonsei University; Hong-Yeop Song*, Yonsei University
 
Remote estimation over channel with state and feedback
Jhelum Chakravorty*, McGill; Aditya Mahajan, McGill
 
Convexity of mutual information along the heat flow
Andre Wibisono*, UW Madison; Varun Jog, UW Madison
 
3:55 On the capacity of private computation
Hua Sun, University of North Texas; Syed Jafar*, UC Irvine
 
On task scheduling in coded computing
Mehmet Aktas, Rutgers; Pei Peng, Rutgers; Emina Soljanin*, Rutgers
 
A novel coding scheme for encoding and iterative soft-decision decoding of binary BCH codes of prime lengths
Shu Lin*, UC Davis; Khaled Abdel-Ghaffar, UC Davis; Juane Li, AppoTech USA Inc.; Keke Liu, Broadcom
 
Self-predicting Boolean functions
Nir Weinberger*, Tel Aviv University; Ofer Shayevitz, Tel Aviv University
 
Circular quadrature amplitude modulations for probabilistic and geometric shaping
Johannes Van Wonterghem, Ghent University, Belgium; Joseph J. Boutros*, Texas A&M Qatar; Marc Moeneclaey, Ghent University, Belgium
 
A Shearer-Type Lemma for Mutual Information
Omri Weinstein*, Columbia; Kasper Green-Larsen, Aarhus University; Huacheng Yu, Harvard
 
4:15 Break
Distributed Storage Systems
Chair: Eirik Rosnes, Simula@UiB
IoT Communication and Localization
Chair: Dinesh Bharadia, UCSD
Polar Codes
Chair: Ruediger Urbanke, EPFL
Biological Systems
Coding for Storage
Chair: Alexandre Graell i Amat, Chalmers
Matrix Completion
Chair: Galen Reeves, Duke
4:30 Explicit MSR codes with optimal access and optimal sub-packetization for d = k+1, k+2, k+3
Myna Vajha, Indian Institute Of Science; Balaji Babu, Indian Institute Of Science; P Vijay Kumar*, IISc and USC
 
Enabling Low-power Computation on Internet-of-Things
Pengyu Zhang*, Stanford
 
Grab a cup of coffee and revisit the scaling exponent of polar codes
Ori Shental*, Bell Labs
 
Synergies between cloud-fog-thing and brain-spinal cord-nerve networks
Eren Balevi, University Of South Florida; Richard Gitlin*, University Of South Florida
 
Constrained codes for reliable flash memories
Akiko Manada*, University Of Electro-Communications Tokyo; Takahiro Ota, Nagano Prefectural Institute of Technology; Hiroyoshi Morita, University Of Electro-Communications Tokyo
 
Spectral aspects of symmetric matrix signings
Alexandra Kolla*, University Of Colorado Boulder; CharlesCar Carlson, University Of Illinois; Karthik Chandrasekaran, University Of Illinois; Hsien Chih Chang, University Of Illinois
 
4:50 Novel latency bounds for distributed coded storage
Parimal Parag, Indian Institute Of Science; Jean-Francois Chamberland*, Texas A&M
 
Battery free cameras with analog backscatter and interactive compression
[Joshua Smith]
 
Fast polarization for processes with memory
Boaz Shuval, Technion; Ido Tal*, Technion
 
Fluctuations of molecule numbers in cells: a queueing model
Justin Dean, University Of Bristol; Ayalvadi Ganesh*, University Of Bristol; Edward Crane, University Of Bristol
 
Code constructions for distributed storage with low repair bandwidth and low repair complexity
Siddhartha Kumar, Simula@UiB; Alexandre Graell i Amat, Chalmers University Of Technology Sweden; Iryna Andriyanova, ENSEA; Fredrik Brannstrom*, Chalmers University Of Technology Sweden; Eirik Rosnes, Simula@UiB
 
Matrix completion for structured observations
Denali Molitor, UC Los Angeles; Deanna Needell*, UC Los Angeles
 
5:10 Unique reconstruction of coded sequences from multiset substring spectra
Ryan Gabrys*, SPAWAR; Olgica Milenkovic, University Of Illinois
 
Battery-Free Wireless Systems: Drone-based Networks, Robotic Manipulation, and Virtual Reality
Yunfei Ma*, MIT; Nicholas Selby, MIT; Fadel Adib, MIT
 
Reduced complexity decoding of polar codes with Reed-Solomon kernel
Peter Trifonov*, St.Petersburg Polytechnic University
 
Open problems in computational immunology
Yana Safonova*, UC San Diego; Pavel Pevzner, UC San Diego
 
Coding for access in memory systems
Ankit Rawat, MIT; Avhishek Chatterjee, IIT Madras; Sriram Vishwanath*, University Of Texas
 
Maximum entropy low-rank matrix completion and recovery
Simon Mak, Georgia Tech; Yao Xie*, Georgia Tech
 

Wednesday

8:15 Continental Breakfast
Information Theory
Chair: Henry Pfister, Duke University
Learning and Recommendation Systems
Chair: Soheil Feizi, Stanford
Ranking and Bandits
Chair: Meisam Razaviyayn, USC
Graphs
Chair: Vincent Gripon, IMT Atlantique
Coding Theory 3
Chair: Soheil Mohajer, University of Minnesota
Estimation and Prediction
Chair: Teemu Roos, University of Helsinki
9:00 Additivity of information in multilayer networks via additive Gaussian noise transforms
Galen Reeves*, Duke
 
Learning the optimal ranking algorithm for search or recommendation results
Charles Elkan*, UC San Diego
 
Regional multi-armed bandits
Cong Shen*, University Of Sci. & Tech. Of China; Zhiyang Wang, University Of Sci. & Tech. Of China; Ruida Zhou, University Of Sci. & Tech. Of China
 
Source location and double metric dimensions of random graphs.
Brunella Spinelli, EPFL; Elisa Celis, EPFL; Patrick Thiran*, EPFL
 
Coded computation against straggling decoders for network function virtualization
Malihe Aliasgari, NJIT; Joerg Kliewer*, NJIT; Osvaldo Simeone, King's College
 
Gravitational allocation for uniform points on the sphere
Nina Holden, MIT; Yuval Peres*, Microsoft; Alex Zhai,
 
9:20 Strong converses are just edge removal properties
Oliver Kosut*, Arizona State; Joerg Kliewer, NJIT
 
Condition number-free query and active learning of linear families
Xue Chen, UT Austin; Eric Price*, UT Austin
 
Ranking from group-wise comparisons
Minje Jang, KAIST; Sunghyun Kim, KAIST; Changho Suh*, KAIST
 
An information theoretic approach to correlated graph matching
Farhad Shirani*, NYU; Siddharth Garg, NYU; Elza Erkip, NYU
 
Cascade determinant codes
Mehran Elyasi, University Of Minnesota; Soheil Mohajer*, University Of Minnesota
 
Online model selection for sequential prediction
Mitas Ray, UC Berkeley; Vidya Muthukumar, UC Berkeley; Anant Sahai*, UC Berkeley
 
9:40 Variational Bayesian Learning for Channel Estimation and Transceiver Determination
Christo Kurisummoottil Thomas, EURECOM; Dirk Slock*, EURECOM
 
Tight regret bounds for a latent variable model of recommendation systems
Mina Karzand*, MIT; Guy Bresler, MIT
 
Opportunistic bandits
Huasen Wu, UC Davis; Xueying Guo, UC Davis; Xin Liu*, UC Davis
 
Lower bounds on the Bayes risk of the Bayesian BTL model with applications to comparison graphs
Mine Alsan*, National University Of Singapore; Ranjitha Prasad ; Vincent Y. F. Tan
 
On deletion/insertion of 0's errors and aymmetric error control codes
Luca Tallini*, University of Teramo - ITALY; Bella Bose, OSU
 
Effective memory shrinkage in estimation
Ayush Jain, UC San Diego; Himanshu Tyagi*, Indian Institute Of Science
 
10:00 Break
10:15 Graduation-Day Talks
Network coding and applications
Chair: Michael Gastpar, EPFL
Security and privacy
Chair: Pierre Moulin, UIUC
Optimization problems
Chair: Maria Florina (Nina) Balcan, CMU
Communication Theory
Chair: Richard Wesel, UCLA
Learning
Chair: Sanjoy Dasgupta, UC San Diego
Statistics: theory and applications
Chair: Charles Elkan, Amazon & UCSD
10:15 Sum-networks from Incidence Structures: Construction and Capacity Analysis
Ardhendu Shekhar Tripathy*, Iowa State; Aditya Ramamoorthy, Iowa State
 
A scalable framework for secure distributed computing
Heecheol Yang*, Seoul National University; Jungwoo Lee, Seoul National University
 
A parallelizable acceleration framework for packing linear programs
Palma London, Caltech; Shai Vardi*, Caltech; Adam Wierman, Caltech; Hanling Yi, CUHK
 
Symmetric vs. Sum Capacity of Rayleigh MAC: Non Asymptotic Analysis
Elad Domanovitz*, Tel Aviv University; Uri Erez, Tel Aviv University
 
Q-Learning with nearest neighbors
Devavrat Shah, MIT; Qiaomin Xie*, MIT
 
Optimal estimation of Gaussian mixtures via denoised method of moments
Yihong Wu, Yale; Pengkun Yang*, University Of Illinois
 
10:35 The explicit coding rate region of symmetrical multilevel diversity coding
Tao Guo*, CUHK; Raymond Yeung, CUHK
 
Robust privacy guarantees for privacy-utility trade-offs
Mario Diaz*, Arizona State; Hao Wang, Harvard; Flavio Calmon, Harvard; Lalitha Sankar, Arizona State
 
Symmetry, saddle points, and global optimization landscape of nonconvex matrix factorization
Xingguo Li*, UMN; Junwei Lu, Princeton; Raman Arora, Johns Hopkins; Jarvis Haupt, UMN; Han Liu, Northwestern; Zhaoran Wang, Northwestern; Tuo Zhao, Georgia Tech
 
A duality between covering and sampling
Jingbo Liu*, Princeton
 
Unraveling the mysteries of stochastic gradient descent on deep neural networks
Pratik Chaudhari*, UC Los Angeles; Stefano Soatto, UC Los Angeles
 
Optimal distribution learning up to permutations
Yanjun Han, Stanford; Jiantao Jiao*, Stanford; Tsachy Weissman, Stanford
 
10:55 Stochastic linear control over binary channels via posterior matching
Anusha Lalitha*, UC San Diego; Anatoly Khina, Caltech; Victoria Kostina, Caltech; Tara Javidi, UC San Diego
 
Dynamic watermarking: Active defense of networked cyber-physical systems
Bharadwaj Satchidanandan*, Texas A&M; P. R. Kumar, Texas A&M
 
Optimal distributed deadline scheduling
Yorie Nakahira*, Caltech; Andres Ferragut ; Adam Wierman
 
Coding structures for multiple-access channel with feedback
Mohsen Heidari Khoozani*, University Of Michigan; Sandeep Pradhan, University Of Michigan
 
What role can machine learning play in communication system design?
Nariman Farsad*, Stanford; Andrea Goldsmith, Stanford
 
On the structure of the least favorable prior distributions
Alex Dytso*, Princeton; Ronit Bustin, Technion; H. Vincent Poor, Princeton; Shlomo Shamai , Technion
 
11:15 Mini Break
11:15 Game theoretic approaches for design of information-centric networks (ICN)
Seyedmohammad Hajimirsadeghi*, WINLAB, Rutgers University; Narayan Mandayam, WINLAB, Rutgers University; Alex Reznik, HPE
 
Private information retrieval over wiretap channel II: Asymmetry, privacy and security.
Karim Banawan*, University Of Maryland College Park; Sennur Ulukus, University Of Maryland College Park
 
Almost condition-free constrained convex optimization
Yen-Huan Li*, École polytechnique fédérale de Lausanne; Carlos Riofrio ; Volkan Cevher, École polytechnique fédérale de Lausanne
 
Coded computing: Mitigating fundamental bottlenecks in large-scale data analytics
Songze Li*, USC; Qian Yu, USC; Mohammad Ali Maddah-Ali, Sharif University Of Technology; Salman Avestimehr, USC
 
Diameter-based active learning
Christopher Tosh*, UC San Diego; Sanjoy Dasgupta, UC San Diego
 
Rank-one NMF-based initialization for NMF and relative error bounds under a geometric assumption
Zhaoqiang Liu*, National University Of Singapore; Vincent Tan, National University Of Singapore
 
11:35 Communications hardware meets information theory and machine learning
Alexios Balatsoukas-Stimming*, EPFL; Andreas Burg, EPFL
 
DARPA spectrum collaboration challenge: The convergence of computation, communication, and collaboration
Dennis Ogbe*, Purdue
 
Repeated sales with multiple strategic buyers
Nicole Immorlica, Microsoft; Brendan Lucier, Microsoft; Emmanouil Pountourakis*, UT Austin; Samuel Taggart, Oberlin College
 
Joint Sporadic Device Activity Detection and Sparse Multipath Massive MIMO Channel Estimation Using Approximate Message Passing
Faouzi Bellili*, University Of Toronto; Wei Yu, University Of Toronto
 
Mutua information Estimation and its Applications to Machine Learning
Shuyang Gao*, University of Southern California; Greg Ver Steeg, University of Southern California; Aram Galstyan, University of Southern California
 
Geometric information divergence: Theory and practice
Salimeh Yasaei Sekeh*, University Of Michigan; Morteza Noshad, University Of Michigan; Kevin Moon, University of Yale; Alfred Hero, University Of Michigan
 
12:05 Graduation-Day Posters and Catered Lunch






12:05 Hypothesis testing under privacy constraints
Jiachun Liao*, Arizona State University; Lalitha Sankar, Arizona State University; Vincent Tan, NUS; Flavio Calmon, Harvard
 
Structured Exploration via Deep Hierarchical Coordination
Stephan Zheng*, Caltech; Yisong Yue, Caltech
 
On a simple characterization of secure exact-repair regenerating codes
Fangwei Ye*, CUHK; Shiqiu Liu, CUHK; Kenneth W. Shum, CUHK; Raymond W. Yeung, CUHK
 
Geometry of sets on the Hamming sphere and binary symmetric relay channel
Xiugang Wu*, Stanford; Leighton Barnes, Stanford; Ayfer Ozgur, Stanford
 
Throughput-optimal scheduling for multi-hop networked transportation systems with switch-over delay
Ping-Chun Hsieh*, Texas A&M; Xi Liu, Texas A&M; Jian Jiao, Texas A&M; I-Hong Hou, Texas A&M; Yunlong Zhang, Texas A&M; P. R. Kumar, Texas A&M
 
Decentralized status update for age-of-information optimization in wireless multiaccess channels
Zhiyuan Jiang*, Tsinghua University; Bhaskar Krishnamachari, University of Southern California; Zhisheng Niu, Tsinghua University
 
12:05 Analysis of fast alternating minimization for structured dictionary learning
Saiprasad Ravishankar, University Of Michigan; Anna Ma*, Claremont Graduate University; Deanna Needell, UC Los Angeles
 
High resolution travel time tomography with local sparsity regularization and dictionary learning
Michael Bianco*, UC San Diego; Peter Gerstoft, UC San Diego
 
Syndrome-coupled rate-compatible error-correcting codes: theory and applications to flash memories
Pengfei Huang*, UC San Diego; Yi Liu, UC San Diego; Xiaojie Zhang, CNEX Labs; Paul H. Siegel, UC San Diego; Erich F. Haratsch, Seagate Technology
 
Pufferfish privacy mechanisms for correlated data
Shuang Song*, UC San Diego; Yizhen Wang, UC San Diego; Kamalika Chaudhuri, UC San Diego
 
Sequential methods for complex systems under transient dynamics
Shaofeng Zou*, UIUC; Georgios Fellouris, UIUC; Venugopal Veeravalli
 
Harnessing Bandit Online Learning for Low-Latency Fog Computing
Tianyi Chen*, University Of Minnesota; Georgios Giannakis, University Of Minnesota
 
12:05 Solving jigsaw puzzle by the graph connection Laplacian
Vahan Huroyan*, University Of Minnesota; Gilad Lerman, University Of Minnesota; Hau-tieng Wu, Duke
 
Low algebraic dimension matrix completion
Greg Ongie*, University Of Michigan; Daniel Pimentel-Alarcon, Georgia State University; Rebecca Willett, University Of Wisconsin; Laura Balzano, University Of Michigan; Robert Nowak, University Of Wisconsin
 
Exploitation of stragglers in synchronous SGD
Nuwan Ferdinand*, University Of Toronto; Stark Draper, University Of Toronto
 
miniRadar: A low power IEEE 802.15.4 transceiver based implementation of bistatic radar
Pradipta Ghosh*, USC; Jenny Xie, Wellesley College; Bhaskar Krishnamachari, USC
 
Entropic Causality
Murat Kocaoglu*, UT Austin; Alexandros G. Dimakis, UT Austin; Sriram Vishwanath, UT Austin; Babak Hassibi, Caltech
 
Coded distributed iterative and inverse problems with applications to graph learning
Yaoqing Yang*, CMU; Pulkit Grover, CMU; Soummya Kar, CMU
 
12:05 Sublinear-time adaptive data analysis
Benjamin Fish*, University Of Illinois Chicago; Lev Reyzin, University Of Illinois Chicago; Benjamin Rubinstein, University Of Melbourne
 
The Gaussian interference channel in the presence of malicious jammers
Fatemeh Hosseinigoki*, Arizona State; Oliver Kosut, Arizona State
 
Edge-assisted on-sensor information selection for bandwidth-constrained wireless systems
Igor Burago*, UC Irvine; Marco Levorato, UC Irvine
 
Emergence of invariance and disentangling in deep representations
Alessandro Achille*, UC Los Angeles; Stefano Soatto, UC Los Angeles
 

 

 
2:00 Plenary: The Emergence Theory of Deep Learning: Perception, Information Theory and PAC Bayes, Stefano Soatto, UCLA and AWS/Amazon AI
2:45 Plenary: From Optimization to Statistical Learning: Two Vignettes from the Interface, Martin Wainwright, UC Berkeley
3:30 Group Photo
3:45 Winter Olympics
6:00 Banquet

Thursday

8:15 Continental Breakfast
Optimization
Chair: Clement Canonne, Stanford
Phase Retrieval
Chair: Yu Xiang, Harvard
High Dimensional Statistics
Chair: Nathan Srebro, TTIC
Social Networks
Chair: Massimo Franceschetti, UC San Diego
Coding Theory
Chair: Marco Mondelli, Stanford
Graph Signal Processing 1
Chair: Geert Leus, TU Delft
9:00 Spectral methods for nonconvex optimization: Sharp asymptotics and phase transitions
Yue Lu*, Harvard
 
Fundamental limits of weak recovery with applications to phase retrieval
Marco Mondelli*, Stanford; Andrea Montanari, Stanford
 
Size-independent sample complexity of neural networks
Noah Golowich, Harvard; Alexander Rakhlin, UPenn; Ohad Shamir*, Weizmann Institute of Science
 
On the effect of information evolution on spreading processes in social networks
Rashad Eletreby, CMU; Yong Zhuang, CMU; Osman Yagan*, CMU
 
Decoding Reed-Muller codes using minimum-weight parity-checks
Mengke Lian, Duke; Elia Santi, Universita Di Parma; Christian Häger, Duke and Chalmers; Henry Pfister*, Duke
 
How to choose a graph Fourier transform
Benjamin Girault, USC; Antonio Ortega*, USC
 
9:20 SamBaTen: Sampling-based batch incremental tensor decomposition
Ekta Gujral, UC Riverside; Ravdeep Pasricha, UC Riverside; Vagelis Papalexakis*, UC Riverside
 
Exact analysis of PhaseMax for phase retrieval
Babak Hassibi*, Caltech; Fariborz Salehi, Caltech; Ehsan Abbasi, Caltech
 
A flexible framework for hypothesis testing in high-dimensions
Adel Javanmard*, University of Southern California (USC
 
What do your friends tell about you? Prediction of individual political participation using social network information
Lorenzo Coviello*, Google; Robert Bond, Ohio State; James Fowler, UC San Diego; Massimo Franceschetti, UC San Diego
 
Two-dimensional algebraic error correcting codes
Shayan Garani*, Indian Institute Of Science Bangalore India; Shounak Roy, Indian Institute Of Science Bangalore India
 
Network inference from graph signals
Santiago Segarra*, MIT; Antonio Marques, King Juan Carlos University; Gonzalo Mateos, University Of Rochester; Alejandro Ribeiro, UPenn
 
9:40 Adaptive sample size methods for empirical risk minimization
Aryan Mokhtari*, MIT
 
Optimization-based AMP for phase retrieval; The impact of initialization and regularization
Arian Maleki*, Columbia; Junjie Ma, Columbia; Ji Xu, Columbia
 
Distributed statistical estimation of high-dimensional and nonparametric distributions with communication constraints
Yanjun Han, Stanford; Pritam Mukherjee, Stanford; Ayfer Ozgur*, Stanford; Tsachy Weissman, Stanford
 
Geographic and temporal trends of fake news consumption during the 2016 US presidential election
Adam Fourney, Microsoft; Miklos Racz, Princeton; Gireeja Ranade*, NEC Labs; Markus Mobius, Microsoft; Eric Horvitz, Microsoft
 
Distributed multi-user secret sharing
Mahdi Soleymani, University Of Michigan; Hessam Mahdavifar*, University Of Michigan
 
An inside look at deep neural networks using graph signal processing
Vincent Gripon*, IMT Atlantique; Antonio Ortega, USC; Benjamin Girault, USC
 
10:00 Coded cooperative computation for Internet of things
Yasaman Keshtkarjahromi, University Of Illinois Chicago; Hulya Seferoglu*, University Of Illinois Chicago
 
Phase retrieval using alternating minimization in a batch setting
Teng Zhang*, University of Central Florida
 
Blessings of dimensionality for sample complexity in latent factor models
Greg Ver Steeg*, USC; Aram Galstyan, USC
 

 
Sidon sets and their applications in coding theory
Mladen Kovacevic, National University Of Singapore; Vincent Tan*, National University Of Singapore
 
Filtering graph signals
Geert Leus*, Delft University Of Technology; Elvin Isufi, Deutsche Telekom; Mario Coutino, Delft University Of Technology; Sundeep Chepuri, Delft University Of Technology
 
10:20 Break
Optimization 2
Chair: Yue Lu, Harvard
Statistical Inference
Chair: Aylin Yener, Penn State
Matrix Factorization
Chair: Shai Vardi, Caltech
Deep Learning
Chair: Vitaly Feldman, Google Brain
Distributed Algorithms and Computation
Chair: Amin Karbasi, Yale
Graph Signal Processing 2
Chair: Peter Gerstoft, UCSD
10:40 Implicit regularization in nonconvex statistical estimation
Cong Ma, Princeton; Kaizheng Wang, Princeton; Yuejie Chi, CMU; Yuxin Chen*, Princeton
 
Teaching and learning in uncertainty
Varun Jog*, University Of Wisconsin
 
Sparse PCA hardness via random rotations
Matthew Brennan, MIT; Guy Bresler*, MIT; Wasim Huleihel, MIT
 
The generalization mystery of deep learning
Sanjeev Arora, Princeton; Rong Ge, Duke; Behnam Neyshabur*, Institute for Advanced Study; Yi Zhang, Princeton
 
Distributed mechanism design for unicast transmission
Nasimeh Heydaribeni, University Of Michigan; Achilleas Anastasopoulos*, University Of Michigan
 
The tensor valued graphical lasso (TeraLasso)
Kristjan Greenewald, Harvard; Shuheng Zhou, University Of Michigan; Alfred Hero*, University Of Michigan
 
11:00 Lazifying conditional gradient descent algorithms
Gábor Braun, Georgia Tech; Sebastian Pokutta*, Georgia Tech; Daniel Zink, Georgia Tech
 
Globally consistent algorithms for mixture-of-experts
Ashok Varadhan, University Of Illinois; Sreeram Kannan*, University Of Washington; Pramod Viswanath, University Of Illinois
 
MEDRoP: Memory-efficient dynamic robust PCA
Praneeth Narayanamurthy, Namrata Vaswani; Namrata Vaswani*, Iowa State
 
Learning ReLUs via gradient descent
Mahdi Soltanolkotabi*, USC
 
Distributed algorithms: Tsitsiklis and beyond
Vivek Borkar*, IIT Bombay; Suhail Mohmad Shah, IIT Bombay
 
The Expxorcist: Nonparametric graphical models via conditional exponential densities
Arun Sai Suggala, CMU; Mladen Kolar, University Of Chicago; Pradeep Ravikumar*, CMU
 
11:20 The minimum Euclidean norm point in a polytope: Wolfe's method is exponential
Jesús De Loera, UC Davis; Jamie Haddock, UC Davis; Luis Rademacher*, UC Davis
 
Spatially fair sampling from big data
Mostafa Rahmani, University of Central Florida; George Atia*, University of Central Florida
 
Nonlinear models for matrix completion
Daniel Pimentel-Alarcon, Georgia State University; Greg Ongie, University Of Michigan; Laura Balzano, University Of Michigan; Rebecca Willett*, University Of Wisconsin; Robert Nowak, University Of Wisconsin
 
Robustness of neural networks against storage media errors
Minghai Qin*, Western Digital Corp.
 
Fast algorithms for distributed optimization
Cesar Uribe, University Of Illinois; Soomin Lee, Facebook; Alexander Gasnikov, MIPT, Moscow, Russia; Angelia Nedich*, Arizona State University
 
Random walks for diffusion of learning functions over graphs
Dimitris Berberidis, University Of Minnesota; Thanos Nikolakopoulos, University Of Minnesota; Georgios B. Giannakis*, University Of Minnesota
 
11:40 On fundamental limitations of stochastic gradient optimization algorithms
Tatiana Tatarenko*, Arizona State
 
Inference task oblivious data reduction
Wenwen Zhao, UC Davis; Shuguang Cui, UC Davis; Lifeng Lai*, UC Davis
 
Separation-free super-resolution from compressed measurements is possible: An orthonormal atomic normal minimization approach
Weiyu Xu*, University of Iowa; Jirong YI, University of Iowa; Soura Dasgupta, University of Iowa; Jian-Feng Cai, University of Iowa; Mathews Jacob, University of Iowa; Myung Cho, University of Iowa
 
On pattern adaptation for the multiscale scan statistic
James Sharpnack*, UC Davis
 
Coded distributed matrix multiplications: Some recent results and open problems
Viveck Cadambe*, Penn State
 
Direction of arrival estimation using dual LASSO with Reed-Solomon error evaluation algorithms
Shrief Rizkalla, TU Wien; Christoph Mecklenbrauker*, TU Wien; Ralf Müller, FAU Erlangen-Nürnberg
 
12:00 Lunch
1:30 Plenary: Going Off the Deep End with Deep Learning, Richard Baraniuk, Rice University
2:15 Plenary: Towards Theoretical Understanding of Deep Learning, Sanjeev Arora, Princeton University
3:00 Break
Statistical Signal Processing
Chair: Babak Hassibi, Caltech
Distribution Learning and Testing
Chair: Yihong Wu, Yale
Structured Learning
Chair: Behrouz Touri, UCSD
Testing and Learning
Chair: Sreeram Kannan, University of Washington
Privacy and Information Retrieval
Chair: Oliver Kosut, Arizona State
Finance/Security
Chair: Lav Varshney, UIUC
3:15 Linear spectral estimation for phase retrieval
Christoph Studer*, Cornell; Ramina Ghods, Cornell; Andrew Lan, Princeton; Tom Goldstein, University Of Maryland
 
On the complexity of learning from label proportions
Benjamin Fish, University Of Illinois Chicago; Lev Reyzin*, University Of Illinois Chicago
 
Learning sequential decision policies with kernel methods
Diyan Teng, Ohio State; Emre Ertin*, Ohio State
 
Dynamic low-rank matrix recovery in blind deconvolution and phase retrieval
Salman Asif*, UC Riverside
 
Privacy while broadcasting in index coding
Mohammed Karmoose, UC Los Angeles; Linqi Song, UC Los Angeles; Martina Cardone*, University Of Minnesota; Christina Fragouli, UC Los Angeles
 
Covert key generation over a classical-quantum channel
Matthieu Bloch*, Georgia Tech; Mehrdad Tahmasbi, Georgia Tech
 
3:35 Radar autofocus using sparse blind deconvolution
Hassan Mansour*, Mitsubishi Electric Research Laboratories; Ulugbek Kamilov, Washington University of St. Louis; Dehong Liu, Mitsubishi Electric Research Laboratories; Petros Boufounos, Mitsubishi Electric Research Laboratories
 
Estimating learnability: approximating signal and noise in the sublinear data regime
Gregory Valiant*, Stanford
 
Hidden integrality and exponential rates of convex relaxations for learning discrete structures
Yudong Chen*, Cornell; Yingjie Fei, Cornell
 
On achieving zero delay with power-of-d-choices load balancing
Xin Liu, Arizona State; Lei Ying*, Arizona State
 
PIR in the time of collusions
Salim El Rouayheb*, Rutgers; Rafael G.L. D'Oliveira, Rutgers
 
Credit bubbles and system memory
Ben Honig ; Michael Honig*, Northwestern; John Morgan, UC Berkeley; Ken Steiglitz, Princeton
 
3:55 Satellite positioning with large constellations
Urs Niesen*, Qualcomm; Olivier Leveque, EPFL
 
Recent advances in algorithmic high-dimensional robust statistics
Ilias Diakonikolas*, USC
 
Learning to control unknown systems
Mukul Gagrani, USC; Rahul Jain*, USC; Ashutosh Nayyar, USC; Yi Ouyang, UC Berkeley
 
Novel impossibility results for group testing
Sidharth Jaggi*, University of Massachusetts at Amherst; Arya Mazumdar, University of Massachusetts at Amherst; Sidharth Jaggi*, CUHK
 
Physical layer secure communications over wireless channels via modulation on zeros
Philipp Walk*, Caltech; Urbashi Mitra, USC; Babak Hassibi, Caltech
 
Blockchains in finance
Rainer Lenz ; Ulrich Tamm*
 
4:15 Break
Submodular Maximization
Chair: Angelia Nedich, Arizona State
Distribution Learning and Testing 2
Chair: Ilias Diakonikolas, USC
Information Theoretic Inference
Chair: Gregory Valiant, Stanford
Sparse Signal Processing
Chair: Christoph Studer, Cornell
Game Theory
Chair: Alfred Hero, University of Michigan
Privacy
Chair: Lalitha Sankar, Arizona State
4:30 Decentralized Submodular Maximization: Bridging Discrete and Continuous Settings
Hamed Hassani*, UPenn
 
Testing conditional independence of discrete distributions
Clement Canonne*, Stanford; Ilias Diakonikolas, USC; Daniel Kane, UC San Diego; Alistair Stewart, USC
 
Learning non-parametric Markov networks with mutual information
Janne Leppä-aho, University Of Helsinki; Santeri Räisänen, University Of Helsinki; Xiao Yang, Apple Inc.; Teemu Roos*, University Of Helsinki
 
Sparse sensing for learning an obstructive propagation environment and optimal aerial relay placement
Junting Chen*, USC; Urbashi Mitra, USC
 
Robust optimistic mirror decent and zero-sum games
Ehsan Asadi Kangarshahi, EPFL; Ya-Ping Hsieh, EPFL; Volkan Cevher*, EPFL
 
Generative adversarial privacy
Peter Kairouz*, Stanford; Chong Huang, Arizona State; Mark Chen, Stanford; Lalitha Sankar, Arizona State; Ram Rajagopal, Stanford
 
4:50 Gradient methods for submodular maximization
Amin Karbasi*, Yale
 
Absolute discounting: One (Categorical) distribution estimator for all dimensions
Moein Falahatgar, UC San Diego; Mesrob Ohannessian*, TTI Chicago; Alon Orlitsky ; Venkatadheeraj Pichapati
 
The computational power of relative entropy
Venkat Chandrasekaran*, Caltech
 
Provably accurate double-sparse coding
Thanh Nguyen, Iowa State; Raymond Wong, Texas A&M; Chinmay Hegde*, Iowa State
 
Quantized mechanisms for allocating divisible resources
Hao Ge, Northwestern; Randall Berry*, Northwestern
 
Data pre-processing for discrimination prevention: Information-theoretic optimization and analysis
Flavio Calmon*, Harvard; Dennis Wei, IBM; Bhanukiran Vinzamuri, IBM; Karthikeyan N. Ramamurthy, IBM; Kush R. Varshney, IBM
 
5:10
 
Communication efficient, sample optimal, linear time locally private learning of discrete distributions
Jayadev Acharya*, Cornell; Ziteng Sun, Cornell; Huanyu Zhang, Cornell
 
Minimum conditional description length
Matt Reyes*, independent consultant; David Neuhoff, University Of Michigan
 
Sparse linear regression under random projection
Shiva Kasiviswanathan*, Amazon AWS AI; Mark Rudelson, University Of Michigan
 
Adversarial contract design for selling data
Parinaz Naghizadeh*, Purdue; Arunesh Sinha, University Of Michigan
 
Maximal alpha-leakage: A tunable measure for information leakage
Jiachun Liao, Arizona State; Oliver Kosut, Arizona State; Lalitha Sankar*, Arizona State; Flavio du Pin Calmon, Harvard
 

Friday

8:15 Continental Breakfast
9:00 Plenary: Measuring Match Outcomes in Online Platforms: The Design of Rating Systems, Ramesh Johari, Stanford University
9:45 Plenary: Beyond Smarts: Toward Correct, Private, Data-Rich Smart Contracts, Ari Juels, Cornell Tech
10:30 Break
Regularization and Generalization in Learning
Chair: Yuxin Chen, Princeton
Distributed Storage 2
Chair: Hessam Mahdavifar, UM Ann Arbor
Networking
Chair: Jia Liu, Iowa State University
Compressed Sensing and Signal Reconstruction
Chair: Junting Chen, USC
Deep Learning and Machine Learning Application
Chair: Sanjeev Arora, Princeton
Deep Neural Nets
Chair: Mahdi Soltanolkotabi, USC
10:45 Implicit regularization in matrix factorization
Suriya Gunasekar*, Toyota Institute Of Technology; Blake Woodworth, Toyota Institute Of Technology; Srinadh Bhojanapalli, Toyota Institute Of Technology; Behnam Neyshabur, Princeton; Nathan Srebro, Toyota Institute Of Technology
 
Distributed storage meets secret sharing on the blockchain
Ravi Kiran Raman, University Of Illinois; Lav Varshney*, University Of Illinois
 
Bonds on coflow scheduling in datacenter networks
Mehrnoosh Shafiee, Columbia; Javad Ghaderi*, Columbia
 
Compression-based snapshot compressive sensing
Shirin Jalali*, Nokia - Bell Labs; Xin Yuan, Nokia - Bell Labs
 
On the detection of atrial fibrillation from ECG signals with deep learning
Matteo Gadaleta, University Of Padova; Steven R. Steinhubl, Scripps; Giorgio Quer*, Scripps
 
Progressive classification and learning: From analytics for deep neural networks to cortex-on-a-chip
John Baras*, University Of Maryland
 
11:05 The implicit bias of logistic regression on separable data
Daniel Soudry, Technion; Elad Hoffer, Technion; Mor Shpigel, Technion; Suriya Gunasekar, Toyota Technological Institute at Chicago; Nathan Srebro*, Toyota Technological Institute at Chicago
 
Mode-suppression: A simple and provably stable chunk-sharing algorithm for P2P networks
Vamseedhar Reddyvari, Texas A&M; Parimal Parag, Indian Institute Of Science; Srinivas Shakkottai*, Texas A&M
 
Optimal sensor scheduling strategies in networked estimation
Marcos Vasconcelos*, USC; Asutosh Nayyar, USC; Urbashi Mitra, USC
 
Oversampled adaptive sensing
Ralf Muller*, FAU Erlangen-Nuernberg; Christoph Mecklenbraeuker, TU Wien; Ali Bereyhi, FAU Erlangen-Nuernberg
 
Second order machine learning
Michael Mahoney*, UC Berkeley
 
Critical points of neural networks: Analytical forms and landscape properties
Yi Zhou, OSU; Yingbin Liang*, OSU
 
11:25 A formal equivalence between implicit and explicit regularization for GLMs
Vince Vu*, Ohio State
 
Codes with locality in the rank and subspace metrics
Swanand Kadhe*, UC Berkeley; Salim El Rouayheb, Rutgers; Iwan Duursma, UIUC; Alex Sprintson, Texas A&M
 
Robustness of flow networks against cascading failures under partial load redistribution
Omur Ozel, CMU; Osman Yagan, CMU; Bruno Sinopoli*, CMU
 
Remote estimation with worst-case estimation guarantee
Mukul Gagrani, Univ. of Southern California; Yi Ouyang, UC Berkeley; Mohammad Rasouli, University Of Michigan; Ashutosh Nayyar*, Univ. of Southern California
 
Communication algorithms via deep learning
Pramod Viswanath*, UIUC
 
Learning neural networks with landscape design
Rong Ge, Duke; Jason Lee*, USC; Tengyu Ma, Facebook
 
11:45 Stability and generalization in adaptive data analysis via calibrated noise addition
Vitaly Feldman*, Google
 
Bridging the gap between the extremes of complete side information versus no side information in consistent distributed storage
Ramy Ali, Nokia; Viveck R. Cadambe, Penn State; Jaime Llora, Bell Labs; Antonia Maria Tulino*, Università Federico II and Bell Labs
 
Jointly optimal routing and caching for arbitrary network topologies
Stratis Ioannidis*, Northeastern University; Edmund Yeh, Northeastern University
 
Signal reconstruction with side information
Cynthia Rush, Columbia; Dror Baron*, NC State
 
5G MIMO data for machine learning: Application to beam-selection using deep learning
Aldebaro Klautau*, Federal Univ. of Para, Brazil; Pedro Batista, Federal Univ. of Para, Brazil; Nuria Gonzalez-Prelcic, University of Vigo; Yuyang Wang, UT Austin; Robert Heath, UT Austin
 
Detecting and correcting for label shift with black box predictors
Zachary Lipton*, CMU; Yuxiang Wang, CMU
 
12:05 Student Posters and light lunch
12:05 Context-aware generative adversarial privacy
Chong Huang*, Arizona State; Peter Kairouz, Stanford; Xiao Chen, Stanford; Lalitha Sankar, Arizona State; Ram Rajagopal, Stanford
 
Edge-Assisted Sensor Control in Health-care IoT
Delaram Amiri*, UC Irvine; Marco Levorato, UC Irvine
 
Latent Laplacian Maximum Entropy Discrimination for Detection of High-Utility Anomalies
Elizabeth Hou*, University Of Michigan; Kumar Sricharan,, Intuit Inc.; Alfred Hero, University Of Michigan
 
A Maximum Entropy Approach to Supervised Learning
Farzan Farnia*, Stanford
 
Memory-efficient kernal PCA via partial matrix sampling and nonconvex optimization: A model-free analysis of local minima
Ji Chen*, University of California, Davis; Xiaodong Li, University of California, Davis
 
Efficient exploration through Bayesian deep Q-networks
Kamyar Azizzadenesheli*, Stanford; Emma Brunskill, Stanford; Animashree Anandkumar, Caltech
 
12:05 Scalable hash-based estimation of divergence measures
Morteza Noshad*, University Of Michigan; Alfred Hero, University Of Michigan
 
Universal image registration and clustering using multivariate information
Ravi Kiran Raman*, University of Illinois, Urbana-Champaign; Lav Varshney, University of Illinois, Urbana-Champaign
 
Improving distributed gradient descent using Reed-Solomon codes
Navid Azizan*, Caltech; Fariborz Salehi, Caltech; Wael Halbawi, Caltech; Babak Hassibi, Caltech
 
Communication Efficient, Sample Optimal, Linear Time Locally Private Learning of Discrete Distributions
Jayadev Acharya, Cornell; Ziteng Sun*, Cornell; Huanyu Zhang, Cornell
 
On the communication complexity over a half-duplex channel.
Ken Hoover, UC San Diego; Russell Impagliazzo, UC San Diego; Alexander Smal, St. Petersburg Department of Steklov Institute of Mathematics of the Russian Academy of Sciences; Ivan Mikhailin*, UC San Diego
 
Online model selection for sequential prediction: From regret to reward
Vidya Muthukumar*, UC Berkeley; Mitas Ray, UC Berkeley; Anant Sahai, UC Berkeley
 
12:05 Erasure coding techniques for faster distributed computing and content download
Sarah E. Anderson, University of St. Thomas; Alyson Fox, Lawrence Livermore National Laboratory; Ann Johnston, Penn State; Gauri Joshi, CMU; Fatemeh Kazemikordasiabi*, Texas A&M; Fiona Knoll, University of Cincinnati; Gretchen L. Matthews, Clemson University; Carolyn Mayer, University Of Nebraska; Emina Soljanin, Rutgers
 
Fast localization of multiple users in mm-wave cells
Esmaeil Karimi*, Texas A&M; Fatemeh Kazemi, Texas A&M; Anoosheh Heidarzadeh, Texas A&M; Alex Sprintson, Texas A&M
 
Maxing and Ranking with Few Assumptions
Venkatadheeraj Pichapati*, UC San Diego
 
Self-organized segregation on the grid
Hamed Omidvar*, UC San Diego; Massimo Franceschetti, UC San Diego
 
The value of information in event triggering: can we beat the data-rate theorem?
Mohammad Javad Khojasteh*, UC San Diego; Jorge Cortés , UC San Diego; Massimo Franceschetti, UC San Diego; Mojtaba Hedayatpour, University Of Regina; Pavankumar Tallapragada, Indian Institute Of Science
 
Improved schemes for asymptotically optimal repair of MDS codes
Ameera Chowdhury*, UC San Diego; Alexander Vardy, UC San Diego
 
12:05 On MIMO Channel Capacity with Output Quantization Constraints
Abbas Khalili Olam*, NYU; Stefano Rini, NCTU; Luca Barletta, Politecnico di Milano; Elza Erkip, NYU; Yonina Eldar, Technion
 
Critical Points of Neural Networks: Analytical Forms and Landscape Properties
[Yi Zhou]
 
[Alireza Khodamoradi]
[Alireza Khodamoradi]
 
[VIPUL GUPTA]
[VIPUL GUPTA]
 

 

 
Detection and Optimization
Chair: Sebastian Pokutta, Georgia Tech
Delay and Networks
Chair: Gireeja Ranade, UC Berkeley
Code Construction and Decoding
Chair: Shayan Garani, IISc
GANs
Chair: Behnam Neyshabur, IAS
Bandits and Ranking
Chair: Michael Mahoney, UC Berkeley
Algorithms
Chair: Jayadev Acharya, Cornell
1:35 Quickest detection of dynamically evolving events in sensor networks
Shaofeng Zou, UIUC; Venu Veeravalli*, UIUC
 
Job delay bounds and asymptotics in cloud computing systems
Weina Wang*, UIUC/ASU; Mor Harchol-Balter, CMU; Haotian Jiang, Tsinghua University; Alan Scheller-Wolf, CMU; R. Srikant, UIUC
 
Information-bit allocation and code construction for polar codes
Jing Jiang, Qualcomm; Gabi Sarkis*, Qualcomm; Yang Yang, Qualcomm; Wei Yang, Qualcomm; Ying Wang, Qualcomm; Joseph Soriaga, Qualcomm
 
Understanding GANs: The LQG setting
Soheil Feizi*, Stanford; Changho Suh, Korea Advanced Institute Of Science; Fei Xia, Stanford; David Tse, Stanford
 
Contextual bandits with stochastic experts
Rajat Sen, UT Austin; Karthikeyan Shanmugam*, IBM T.J Watson; Sanjay Shakkottai, UT Austin
 
Connected dominating sets: new results on an ancient problem
Samir Khuller*, University Of Maryland College Park; Sheng Yang, University Of Maryland College Park
 
1:55 Machine learning and algorithmic techniques for error correction
Anxiao Jiang*, Andrew
 
A non-monetary mechanism for optimal rate control through efficient delay allocation
Tao Zhao, Texas A&M; Korok Ray, Texas A&M; I-Hong Hou*, Texas A&M
 
Blind detection of polar codes
Pascal Giard, EPFL; Alexios Balatsoukas-Stimming, EPFL; Andreas Burg*, EPFL
 
The power of two samples in generative adversarial networks
Zinan Lin, CMU; Ashish Khetan, UIUC; Giulia Fanti, CMU; Sewoong Oh*, UIUC
 
Medoids in almost linear time vis multi-armed bandits
Vivek Bagaria, Stanford; Govinda Kamath, Stanford; Vasilis Ntranos, UC Berkeley; Martin Zhang, Stanford; David Tse*, Stanford
 
Recovering a hidden hamiltonian cycle via linear programming
Vivek Bagaria, Stanford; Jian Ding, UPenn; David Tse, Stanford; Yihong Wu*, Yale; Jiaming Xu, Purdue
 
2:15 Gaussian Mixture Models for Detection of Adversarial Inputs
Pierre Moulin*, UIUC
 
Age of information over networks of preemptive servers
Roy Yates*, Rutgers
 

 
Discrete Wasserstein generative adversarial networks
Rizal Fathony, Technicolor; Naveen Goela*, Technicolor
 
Approximate ranking from pairwise comparisons
Reinhard Heckel*, Rice; Max Simchowitz, UC Berkeley; Kannan Ramchandran, UC Berkeley; Martin Wainwright, UC Berkeley
 
LSH breaks computational chicken-and-egg loop in adaptive sampling and estimations
Anshumali Shrivastava*, Rice
 
2:35 Break
Applications of Neural Network
Chair: Zachary Lipton, CMU
Life Science Applications
Chair: Yana Safonova, UCSD
Optimization and Energy Efficiency
Chair: David Mitchell, New Mexico State
Communications
Chair: Ryan Gabrys, SSC Pacific
Recovery Algorithms and Online Learning
Chair: Yin Sun, Auburn University
Networks
Chair: Weina Wang, UIUC & Arizona State
2:50 Characterizing sparse connectivity patterns in neural networks
Sourya Dey, USC; Kuan-Wen Huang, USC; Peter Beerel, USC; Keith Chugg*, USC
 
Conversations with chemicals: Memory and computing without long oligomers
Chris Rose*, Brown University
 
Smart routing in smart grids
Rasoul Etesami*, University of Illinois at Urbana-Champaign; Walid Saad, Virginia Tech; Narayan Mandayam, WINLAB; Vincent Poor, Princeton University
 
Error-free communication over state-dependent channels with variable-length feedback
Carol Wang*, NUS; Mladen Kovačević, NUS; Vincent Tan, NUS
 
Phase retrieval from local measurements: deterministic measurement constructions and efficient recovery algorithms
Mark Iwen, Michigan State; Brian Preskitt, UC San Diego; Rayan Saab, UC San Diego; Aditya Viswanathan*, University of Michigan - Dearborn
 
High-order momentum: Improving latency and convergence for wireless network optimization
Jia Liu*, Iowa State
 
3:10 Compressing the Internet: Numerical representation of Internet hosts
Armin Sarabi, University Of Michigan; Mingyan Liu*, University Of Michigan
 
Evolution of k-mer frequencies in stochastic mutation systems
Hao Lou, University Of Virginia; Moshe Schwartz, Ben Gurion University; Farzad Farnoud*, University Of Virginia
 
Solar energy sharing in net metered community microgrids: can the social goals be achieved?
Yue Zhao*, Stony Brook University
 
Proactive content caching with improving predictions over time
John Tadrous*, Gonzaga University; Atilla Eryilmaz, Ohio State
 
Maximum likelihood decoding by guessing noise
Muriel Medard*, MIT; Ken Duffy, Hamilton Institute NUI Maynooth
 
Interdependent random geometric networks
Jianan Zhang, MIT; Edmund Yeh*, Northeastern University; Eytan Modiano, MIT
 
3:30 On the information in spike timing: neural codes derived from polychronous groups
Zhinus Marzi, UC Santa Barbara; Joao Hespanha, UC Santa Barbara; Upamanyu Madhow*, UC Santa Barbara; Bruno Olshausen, UC Berkeley
 
Understanding adverse conditions of pregnancy via statistical learning
Ivana Maric*, Stanford; Virginia Winn, Stanford; David Stevenson, Stanford; Gary Shaw, Stanford
 
A primal-dual method for non-smooth composite and distributed optimization
Sepideh Hassan-Moghaddam, USC; Mihailo Jovanovic*, USC
 
Beamforming training for IEEE 802.11ay millimeter wave systems
Claudio da Silva*, Intel; Jonathan Kosloff, Intel; Cheng Chen, Intel; Artyom Lomayev, Intel; Carlos Cordeiro, Intel
 
Integrating online learning and adaptive control in queueing systems with uncertain payoffs
Wei-Kang Hsu, Purdue; Jiaming Xu*, Purdue; Xiaojun Lin, Purdue; Mark Bell, Purdue
 
A dense IoT architecture based on mobile gateways
Francois Baccelli*, UT Austin; Chang Sik Choi, UT Austin; Gustavo De Veciana, UT Austin
 
3:50 Closing Ceremony and traditional piata event by the lawn, followed by a farewell bash with food and drinks at the Gardenia Suite, Rm 200.