Feedback

Snehanshu Saha

Professor, Department of Computer Science and Information Systems & Co-ordinator-APPCAIR

Data Science and Artificial Intelligence
Department of Computer Science and Information Systems,
BITS Pilani K K Birla Goa Campus, Zuarinagar

About the Faculty

Snehanshu Saha
Professor, Computer Science and Information Systems and Co-ordinator-APPCAIR
BITS PILANI K K K BIRLA GOA CAMPUS (JOINED: 31/12/2019)
https://www.bits-pilani.ac.in/goa/appcair/people
https://github.com/sahamath/sym-netv1 (Codes/Data)
Google-Scholar: https://scholar.google.com/citations?user=C-Qm2LcAAAAJ&hl=en (Full List of Publications)
Citations: 1405; h-index: 17; i-10 index: 31
Email: snehanshus@goa.bits-pilani.ac.in; snehanshu.saha@ieee.org; T +91 08322580855
Citations (As BITS PIlani Faculty): 131
Co-Founder and Director of Research-HappyMonk AI Labs (Bangalore and Goa); SM-IEEE, SM-ACM

Academic Training

  • PhD in Shallow water Waves and Partial Differential Equations , Mathematical Sciences - University of Texas Arlington, 2008, Texas, USA.
  • MS Computational Mathematics - Clemson University, , Clemson, SC, USA. 2003
  • BE (Computer Science and Engineering)-Jawharlal Nehru National College of Engineering (First Class with Distinction), 1999
 
I develop methods in Computational Learning Theory (COLT), Mathematics of Data Science (MDS), AstroInformatics (AIM) and novel ML/Modeling applications (AML). These and interpretable Deep learning techniques and acceleration in deep and wide neural networks, aided by chaos-causality paradigm are my current and future research interests. My Open Source Educational Initiatives include the following:
 
1. Machine Learning Blog: https://beginningwithml.wordpress.com/
2. Youtube Channel: https://tinyurl.com/yyg7skpf

 

I have contributed more than 100 articles in top ranked conferences and Journals including several IEEE and ACM Transactions and am a Senior Member-IEEE, Senior Member-ACM and a Fellow-IETE. I contributed to the theory of machine learning/deep learning and meta-heuristics and developed several state-of-the-art methods in

Deep Learning based classification tasks, optimizers for non-convex problems, proposed a novel anomaly detection method for univariate and multivariate time series data and interpretable tools in Deep learning. I have 03 European Patents, several Sponsored Projects and contributed to Open Source Educational Initiatives. I have designed and introduced a new course, Computational Learning Theory for UG/PG students. I led IEEE Computer Society Bangalore chapter to new heights and helped secure them the best global chapter in 2019. I'm  a co-founder of HappyMonk AI, an AI product company.

Note to students: Since I work in foundational machine learning, it is exceedingly difficult for me to accept semester-long projects/SOPs etc. If you're committed for the grind, spanning at least a year, let's talk!

Research Interest

I develop methods in Computational Learning Theory (COLT) and Mathematics of Data Science (MDS) Techniques

 
I focus on: Pointwise smooth approximations of gradient, discovering activation functions from data, smoothness and convexity studies of loss functions, methods in interpretable deep learning, novel loss functions in back-propagation, Chaos-causality in Deep Neural Networks, pruning/compression in Deep Neural Networks, Kernels for Classical Learning
 
Goals: Efficient, reliable and robust Inferences (One Landscape)
 
Dreams: Porting methodological elegance of classical machine learning to deep learning
 
Tools I use: Measure Theory, NonLinear Optimization, Functional Analysis, Econometrics, Dynamical Systems, Algorithms
 
Applications: Biology, Healthcare, AstroInformatics, Transportation, Industrial Systems and Automation
 
I have initiated a new focus group on Astroinformatics, in collaboration with a few colleagues in Indian Institute of Astrophysics and National Institute of Advanced Studies. 
 
All code bases related to the group's research are available at : https://github.com/sahamath
 

Administrative Responsibilities

Department-Level: DAC member: 2020-22; DRC member: 2020-24

Institute-Level: Chair- the Scientometrics and Ranking committee (2022-current); Member-Convocation committee (2022)

University-Level: Core Committee: Website revamping (2022-current); Core Committee: Ranking (2022-current)

New program Design: Computational Economics (2022)