ML Classroom: Classification (pt 1)

Talk description :
This talk will be a gentle introduction to two standard classification techniques - logistic regression, and support vector machines. For each classifier, we will formalize the problem, formulate an objective function, and discuss how to optimize it. Using different data sets, we will compare how the two classifiers stack up to each other in practical settings. The role of L1 and L2 regularization in machine learning will also be discussed. No prior knowledge of any machine learning or statistics is assumed in this talk. 

Speaker bio
Kriti Puniyani is a graduate student at Carnegie Mellon University, working in machine learning applied to computational biology and social media analysis. Kriti received her Masters degree from the Indian Institute of Technology Bombay, where her thesis was in informational retrieval. She has done internships in IBM India Research Lab, and in the Yahoo! technology and research group.