Deep Learning Workshop

Title: Workshop on Deep learning for novice

Abstract
This workshop seeks to deliver fundamentals of machine learning, with a focus on deep learning and the terminologies, methods involved in developing deep nets. It tries to provide practical exposure to build, train and test Deep net models covering a few real-world applications

Session plan:
Fore noon

  • Introduction to Deep Learning: neural networks, representation, backpropagation and Deep models
  • Determining a model capacity and model generalization,
  • Regularization strategies to reduce overfit,
  • Model optimization
Afternoon
  • Hands-on First Neural net and deep learning framework using Keras.
  • Convolution neural nets basics, design of CNN architectures, transfer learning and hands-on CNN.
Speaker

Dr. B. Rajesh Kanna
Associate Professor
School of Computer Science and Engineering
Vellore Institute of Technology

Profile of Speaker
Rajesh Kanna Baskaran is an Associate Professor, School of Computing Science and Engineering, Vellore Institute of Technology (VIT), Chennai, India since 2012, His research interests include digital image analysis, hypergraph based models of images. He currently explores the incorporation of Deep neural networks in microbial analysis to curb the accelerated progress against the global burden of malaria and tuberculosis. His doctoral dissertation "Development of Hypergraph based Models for Selected Image Engineering Applications has proposed two innovative models for contour-based image analysis and proved the efficiency of the models through experiments. ACM SIGSOFT has chosen him as one of the researchers to receive a travel scholarship to attend ACM A.M. Turing Centenary celebration held at San Francisco, California, during June 15-16, 2012. His work on parallelism in noise removal won the Best poster award at 22nd International Symposium on High-Performance Parallel & Distributed Computing, June 17-21, 2013 at New York. He has co-authored many research publications, book chapters and a book in reputed Journals and Conferences. For more details follow his research citations
http://www.scopus.com/authid/detail.url?authorId=25646228400