**Email:** name

at ece

dot ucr

dot edu (name=oymak)

## Welcome

I am an Assistant Professor in the Department of Electrical and Computer Engineering at UC Riverside. Previously, I worked at The Voleon Group and Google and spent time at Berkeley as a postdoctoral scholar at the

Simons Institute.
I received my PhD degree in Electrical Engineering from

Caltech.

I am broadly interested in optimization, statistics, and machine learning. This includes convex/nonconvex optimization problems, nonlinear models, time-series analysis, and large scale distributed algorithms. I am particularly interested in developing principled algorithms with solid theoretical foundation that have good tradeoffs between speed, accuracy, scalability, and the data-efficiency.

Here is my

CV. Some recent publications are listed below.

### Recent papers

●

**Learning Compact Neural Nets:** Proper regularization is critical for speeding up training, improving generalization performance, and learning cost efficient models. In this work, we propose and analyze regularized gradient descent algorithms for learning shallow networks. We introduce covering dimension and show that problem becomes well conditioned and local linear convergence occurs once the amount of data exceeds the covering dimension. We also establish generalization bounds and consider convolutional and sparsity constraints as applications.

*S. Oymak*, ``Learning Compact Neural Networks with Regularization,'' to appear at ICML 2018.

●

**End-to-end Deep Learning:** We propose data-efficient tensor decomposition algorithms for training convolutional neural nets (CNN). We rigorously establish a connection between low-rank tensors and CNNs and propose a multistage approach that can learn convolutional kernels of all layers simultaneously.

*S. Oymak* and M. Soltanolkotabi ``End-to-end Learning of a Convolutional Neural Network via Deep Tensor Decomposition,'' preprint May 2018.

●

**Sharp Time-Data Tradeoffs for Linear Inverse Problems:** We present a unified convergence analysis of the gradient projection algorithm applied to inverse problems. We *very* accurately
characterize the convergence rate associated with a wide variety of random measurement ensembles in terms of the data amount and structural complexity of the model.

*S. Oymak*, B. Recht, and M. Soltanolkotabi, ``Sharp Time-Data Tradeoffs for Linear Inverse Problems,'' Trans. on Info. Theory June 2018.

●

**Learning Feature Nonlinearities with Binned Regression:** What to do when the dependence betweens feature and response is nonlinear? We propose a principled algorithm to learn additive models based on feature binning. Borrowing techniques from high-dimensional statistics, we show that such models can be learned with linear convergence and using near optimal amount of data. Our findings naturally highlight the role of model complexity.

*S. Oymak*, J. Chen, and M. Mahdavi, ``Learning Feature Nonlinearities with Non-Convex Regularized Binned Regression,'' preprint 2017.

●

**Sketching any set with RIP matrices:** We show that for the purposes of dimensionality reduction certain class of structured random matrices behave similarly to random Gaussian matrices. This class includes several matrices for which matrix-vector multiply can be computed in log-linear time, providing efficient dimensionality reduction of general sets.

*S. Oymak*, B. Recht, and M. Soltanolkotabi, ``Isometric sketching of any set via the Restricted Isometry Property,'' Information & Inference, March 2018.

●

**Universality Laws for Randomized Dimension Reduction:** Gaussianity assumption plays a central role in various data science problems. In this work, we show that a large class of random matrices behaves in a identical fashion to Gaussians for various optimization and learning problems.

*S. Oymak* and J. Tropp, ``Universality Laws for Randomized Dimension Reduction, with Applications,'' Information & Inference 2017.

●

**Fast and Reliable Estimation from Nonlinear Observations:** We show that it is possible to quickly learn nonlinear models even if the relation between input and output (i.e. link function) is not known. The key idea is linear approximation of the nonlinearity.

●

**Sample Optimal Bounds for Fast Binary Embedding:** A common strategy for fast document and image retrieval is to create a binary signature of the data by embedding it into Hamming cube. Ideally, we want a fast and space efficient embedding. We show that subsampled Fourier transform indeed provides an almost-optimal embedding strategy even after {+1, -1} quantization.

*S. Oymak*, ``Near-Optimal Sample Complexity Bounds for Circulant Binary Embedding
,'' preliminary results published at ICASSP 2017.

●

**Simultaneously structured models:** In several applications, the model of interest has several structures at the same time. Examples include, quadratic compressed sensing, sparse PCA, low-rank tensor completion and estimating sparse and smooth signals (fused lasso). We show a weakness result for multi-objective convex penalizations that aims to recover and/or estimate such signals. The recently updated paper has simpler failure conditions that apply to a wide-range of measurement ensembles under much less assumptions.

*S. Oymak*, A. Jalali, M. Fazel, Y. C. Eldar, and B. Hassibi, ``Simultaneously Structured Models with Application to Sparse and Low-rank Matrices,'' Trans. on Info. Theory, PDF.

●

**Graph clustering with missing data:** Graph clustering problem arises in community detection, social networking, and complex networks. Given access to noisy and incomplete connectivity information of the graph, can we identify the communities that matter (i.e. clusters)? We develop simple but precise performance bounds for convex optimization based clustering techniques. These bounds reveal intriguing phase transitions as a function of model parameters.

- R. Korlakai Vinayak,
*S. Oymak*, and B. Hassibi, ``Graph Clustering With Missing Data: Convex Algorithms and Analysis,'' NIPS.

My thesis can be found

here. Finally, you can click

here for some personal facts.