Assistant Professor, Knight Foundation School of Computing and Information Sciences, Florida International University
Farhad Shirani is an Assistant Professor at the Knight Foundation School of Computing and Information Sciences at Florida International University. He received his B.Sc. degree in Electrical Engineering from Sharif University of Technology, and M.Sc. degree in Applied Mathematics and M.Sc. and Ph.D. degrees in Electrical Engineering at the University of Michigan at Ann Arbor. He served as a Lecturer and Postdoctoral Research Fellow at University of Michigan in 2017. He was a Research Assistant Professor at New York University in 2017-2020. He was an Assistant Professor at the Electrical and Computer Engineering Department at North Dakota State University 2020-2022. His research interests include privacy, wireless communications, information theory, and machine learning.
Binary classifiers appear in various applications such as learning, estimation, data compression, and hypothesis testing. A binary classifier computes a Boolean function of its input sequence. In this talk, we introduce the concept of the `effective input length’ (EIL) of a binary classifier. The EIL is a measure of how much each of the elements of the input sequence contribute to the classification process. We consider distributed pairs of classifiers operating on correlated sequences of random variables and characterize a fundamental tradeoff between their probability of disagreement, and their respective EILs. We use the characterization to investigate problems in distributed data storage, feature selection and function approximation.