Avatar

Ganga Meghanath

Data Scientist II @ Microsoft

Hello! I’m a Data Scientist at Microsoft part of the Experimentation for Windows (EFW) team. My work focuses on understanding how different features impact user experiences through causal analysis, including both causal discovery and causal inference. I also collaborate with various teams at Microsoft and our customers to provide insights and solutions in causal analysis.

I hold a Master’s degree in Electrical and Computer Engineering from Purdue University, where I researched neural population dynamics and speech decoding under Prof. Joseph Makin. During my time there, I also taught Python programming as a Graduate Teaching Assistant. I earned my Bachelor’s degree in Electrical Engineering from the Indian Institute of Technology, Madras.

Previously, I worked as a Data Scientist in the Bing Ads team at Microsoft, developing machine learning pipelines to detect and reject editorial policy violations and identify threats like fraud and phishing in text ads.

Interests:

Causal Analysis, Machine Learning, Natural Language Processing

Publications

Inferring population dynamics in macaque cortex

Journal of Neural Engineering, 2023

ApproxNet: Content and Contention-Aware Video Analytics System for Embedded Clients

ACM Transactions on Sensor Networks, 2021

Articles

Beginner’s guide to causal discovery: The what, the why, and the how

Published in Data Science at Microsoft, 2024

Causal analysis overview: Causal inference versus experimentation versus causal discovery

Published in Data Science at Microsoft, 2024

Projects

Identify frames in the Tweets of US politicians

Framing refers to wording your opinion on a certain subject to emphasize certain aspects of the topic over the others.

Modeling Ecological Populations - Game Theory

Study the population convergence of N-player game using learned strategies.

Improving robustness of neural networks against adversarial attacks

Study Adversarial attacks and Defence techniques for Machine Learning models.

Memory based Multi-tasking A3C Agent

A memory-incorporated RL framework that can learn to do Multiple tasks through active learning, and effectively reduce catastrophic …

A Hierarchical Approach to Multi Tasking in Reinforcement Learning

Study and evaluation of the performance Hierarchical Reinforcement Learning frameworks in multi-tasking domains using active sampling.

Weather data summarizer using encoder-decoder networks

Implemented a table summarizer for structured weather data using an encoder-decoder model comprising of an attention layer over a …

Word embeddings for native languages

Created a corpus and embeddings for nearly 15 million words in the Indian native language Malayalam.

Team Anveshak, University Rover Challenge, Utah

We build a remote operated all-terrain rover, complete with a robotic manipulator and digger, with an in-built autonomous navigation …

Academic Courses

Mathematical Foundations:

  • MA1101: Calculus I : Functions of One Variable
  • MA1102: Calculus II : Functions of Several Variables
  • MA2040: Probability, Statistics and Stochastic Processes
  • MA2031: Linear Algebra for Engineers
  • EE3110: Probability Foundations for Electrical Engineers
  • Purdue University
  • CS 57700 : Natural Language Processing
  • ECE 69500: Inference & Learning in Generative Models
  • ECE 59500: Introduction to Data Mining
  • ECE 60800: Computational Models and Methods

  • IIT Madras
  • CS7015: Deep Learning
  • CS6700: Reinforcement Learning
  • CS4011: Principles of Machine Learning
  • CS7011: Topics in Reinforcement Learning
  • EE6418: Game Theory
  • EE3004: Control Engineering
  • EE2004: Digital Signal Processing
  • EE4371: Data Structures and Algorithms
  • ID6040: Introduction to Robotics
  • ID7123: Machine Intelligence and Brain Research
  • BT6270: Computational Neuroscience
  • MS3510: Operations Research

Contact