SDSC Special Events

The mini-course will meet on Monday, April 1 and Wednesday, April 3rd from 1:30-3:00pm

This mini-course deals with a circle of problems related to estimation of real valued functionals of high-dimensional and infinite-dimensional parameters of statistical models.

In such problems, it is of interest to estimate one-dimensional features of a high-dimensional parameter represented by nonlinear functionals of certain degree of smoothness defined on the parameter space. The functionals of interest could be often estimated with faster convergence rates than the whole parameter (sometimes, even with parametric rates).

We will discuss some mathematical methods providing a way to develop estimators of functionals of high-dimensional parameters with optimal error rates in classes of functionals of some Hoelder smoothness and even to provide their efficient estimation with parametric rates when the smoothness is sufficiently large.

The main focus will be on functionals of unknown covariance operators in high-dimensional and infinite-dimensional Gaussian models, where the functionals of interest often capture their spectral properties. In particular, we will discuss the role of higher order bias reduction methods and concentration inequalities in these problems.

Vladimir Koltchinskii is a professor of Mathematics at the Georgia Institute of Technology.

Abstract: We propose an alternative paradigm for the Internet of Things (IoT) where machine learning algorithms run locally on severely resource-constrained edge and endpoint devices without necessarily needing cloud connectivity. This enables many scenarios beyond the pale of the traditional paradigm including low-latency brain implants, precision agriculture on disconnected farms, privacy-preserving smart spectacles, etc.

Towards this end, we develop novel tree and kNN based algorithm, called Bonsai and ProtoNN, for efficient prediction on IoT devices — such as those based on the Arduino Uno board having an 8 bit ATmega328P microcontroller operating at 16 MHz with no native floating point support, 2 KB RAM and 32 KB read-only flash memory. Experimental results on multiple benchmark datasets demonstrate that Bonsai and ProtoNN can make predictions in milliseconds even on slow microcontrollers, can fit in KB of memory, have lower battery consumption than all other algorithms while achieving prediction accuracies that can be as much as 30% higher than state-of-the-art methods for resource-efficient machine learning.

Time permitting, I will also discuss our recent results about deploying RNNs on similar sized tiny devices.

Joint work with Manik Varma, Harsha Simhadri, Arun Suggala, Ankit Goyal, Chirag Gupta, Don Dennis, Aditya Kusupati, Shishir Patil, Ashish Kumar.

Biography: I am a member of the Machine Learning and Optimization and the Algorithms and Data Sciences Group at Microsoft Research, Bangalore, India. My research interests are in machine learning, non-convex optimization, high-dimensional statistics, and optimization algorithms in general. I am also interested in applications of machine learning to privacy, computer vision, text mining and natural language processing.
Earlier, I completed my PhD at the University of Texas at Austin under Prof. Inderjit S. Dhillon.