Expert Talks (MLSP)

Here’s a summary of each Expert Talk in the course:

Framework to Manage End-to-End Machine Learning Projects by Kiran R

Learn how to manage machine learning projects from the concept stage to the final completion stage. Various aspects of a typical machine learning project, including how to successfully manage one from scratch will be discussed along with tips on model building and last mile optimization.

Patterns of Insights by Anand S

Once you’ve got the data, the next questions to ask are:

●    How do you get interesting stories out of this data?

●    And how do you narrate these stories?

The first step here is to figure out what the interesting insight is. A fair bit of this is understanding the domain well enough, and people tend to miss out the soft skills aspect. The second step involves converting these insights into a series of questions that the data can answer.

Are there patterns of questions that we can pose to the data and is there a systematic and structured way by which we can explore it? This talk by Anand S will answer these questions and more.

Evaluating Machine Learning Models for Bias by Rajesh Jeyapaul and Prateek Goyal

Although machine learning, by its very nature, is always a form of statistical discrimination, the discrimination becomes objectionable when it places certain privileged groups at a systematic advantage and certain unprivileged groups at a systematic disadvantage. In this hack session, Rajesh and Prateek will discuss the concepts and capabilities of a model to test for biases and explanations along with the wider spectrum of explainability methods, notably data explanations, metrics and persona specific explanations.

Building an End-to-End Credit Risk Model by Arihant Jain

This talk is about developing a Machine Learning model that determines which loan applicants are credible and developing a monitoring framework and feedback loop for the Credit Risk Model.

Hyperparameter Optimization using Bayesian Optimization by Abhishek Periwal

Hyperparameter tuning is an essential part of any machine learning pipeline. With better compute we now have the power to explore more range of hyperparameters quickly but especially for more complex algorithms, the space for hyperparameters remains vast and techniques such as Bayesian Optimization might help in making the tuning process faster. Witness how you can use HyperOpt library uses Bayesian techniques to find the best hyperparameters with a real dataset.

Performing Feature Selection in High-Dimensional Spaces by Sidharth Kumar

Typical feature reduction algorithms suffer from problems relating to class of features (categorical, continuous) and the complexity resulting from the number of features, thus making redundant feature elimination a difficult task. Additionally, it is not clear as to how the objective function affects the feature selection. In this hack session, Sidharth will discuss how to use moving set of thresholds for feature selection, which are optimized based on a Markov Chain Monte Carlo that adjusts the thresholds based on the final model fitness.

Introduction to the World of Reinforcement Learning by Professor Balaraman Ravindran

In this talk, Professor B. Ravindran will introduce you to the crux of what reinforcement learning is and why it’s a distinct learning paradigm as compared to the traditional machine learning approaches. He will also cover the different kinds of applications where reinforcement learning can be applied, and the challenges industries are facing in applying RL.

State of Transfer Learning in NLP by Sudalai Rajkumar (SRK)

NLP's latest pre-trained language models like BERT, GPT2, TransformerXL, XLM, etc. are achieving state of the art results in a wide range of NLP tasks. In this hack session, community favourite and Kaggle Grandmaster SRK will compare the performance of these different pre-trained models along with pre-trained word vector models on classification tasks.

Effective Feature Engineering for Building Better ML Models by Dipanjan Sarkar

In this hack session, Dipanjan will be taking a structured and comprehensive hands-on approach (with Python) to feature engineering, where we will explore two interesting case studies based on real-world problems!

Automating the Machine Learning Pipeline by Dr. Sunil Kumar Chhinnamgari

This conference talk attempts to introduce the concept of AutoML using two real-world case studies.