Machine Learning for Business Processes

On 28th August 2020, my alma-mater, IIIT Naya Raipur (IIITNR) organised 3rd Annual Industry Academia Meet (IAM). This year, they introduced a special session on Carrer Guidance from Alumni for the current students. I feel honoured for being invited to talk about my research and guide other students to make career in Machine Learning. In this blog I’ll give a brief overview of the talk and share the slides.

IAM Poster

IAM at IIITNR is a great intiative started in 2017 where different personalities from IT Indsustry are invited to interact with the students of IIITNR. They deliver special lectures, talk about their work in the industry, judge different student competitions, and most importantly - guide the students on being industry-ready.

Introduction

Since, this event was all about making students industry-ready and guiding them for their career, decided to choose something related to real-world Machine Learning. The topic of my talk was Machine Learning (ML) for Business Processes (BP). There are many reasons for choosing this topic: first, ML has reached that stage where models are being deployed in real world, but they still face problems related to labelled data which is expensive to get’; second, this is what I worked on during my final semester internship at RGU and am working in my PhD.

About the Talk

In this talk, I motivated the students to focus on efficient ways of gathering data in ML ecosystem before going into the learning algorithms. I also gave an example on how to use one such method, Active Learning (AL), for a classification problem. For a quick intro, AL is a method where learning algorithms are allowed to choose the data from which they learn, hence achieving greater accuracy with fewer training labels available.

In this digital-era, although the demand for Data Scientists is high, we can’t ignore the fact that supply is also not low. Thus, I wanted to draw students’ attention towards working on real-world ML ecosystem where these algorithms are required to learn better in a low resource setting as well. That’s why, having some experience in Low Resoucre ML is always going to set them aside from the crowd.

Acknowledgement

First of all, I would like to thank the whole IIITNR family for giving me this wonderful opportunity. I would especially like to thank Vice-Chancellor & Director, Dr PK Sinha for his constant motivation and encouragement. He has always tried to get the best out of his students. I would like to thank the Alumni Association of IIITNR, especially Dr Santosh Kumar and Dr PP Paltani for organising such a wonderful event, that too in online mode. I would also like to thank Dr Muneendra Ojha for always guiding me and motivating me to present my work in this event.

I wish the students of IIITNR very best of luck in their journey. Here are the slides from my talk: