Abstract: Over the years we have seen the evolution of thumb rules to choose the Machine Learning style for a problem - supervised classification vs. regression problems, vs. unsupervised clustering problems. But we will challenge typical formulations and their assumptions. Taking several examples from e-commerce and BFSI domains, we will discuss how a problem can be solved using multiple ML formulations which can also be blended to improve results. Additionally, the dependent variables and the set of features for the Machine Learning model are choices that significantly impact the performance of the machine learning model. Finally, we will discuss how for a machine learning solution to create maximum impact, the model objective should be aligned with the business objective.

About the Speaker: Rahim brings extensive experience in E-commerce and solving challenging business problems using the latest in Machine Learning and Artificial Intelligence. Author of "Data Science for Marketing Analytics" and "The Deep Learning Workshop", Rahim is also a visiting faculty at NMIMS teaching Advanced Machine Learning, Business Problem Solving courses. Additionally, he is a subject matter expect for the various major online learning platforms on these topics. Rahim currently leads the Product Analytics charter at Zalando Marketing Services, Germany.

Register in advance for this webinar