Lokesh Nagalapatti

Research Scholar with CSE Dept, IITB.

IIT Bombay

Hi! I am a Third Year PhD studnet with IIT Bombay, India. I joined PhD in the year 2021 and am fortunate to be advised by Prof. Sunita Sarawagi and Prof. Abir De. I work in the areas of Causal Inference and Algorithmic Recourse. My PhD is generously funded by the Prime Minister Research Fellowship.

Prior to IITB, I spent about 1.5 years with IBM Research working on cutting edge topics like Federated learning, assessing data quality metrics, aggregating user preferences etc. Prior to that, I completed masters from IISc Bangalore. I worked with Prof. M Narasimha Murty on finding outlier robust representations of nodes in Social Networks.

Interests

ML
Causal Inference
Algorithmic Recourse

Education

PhD in Computer Science and Engineering, 2021 - Present

IIT Bombay
M.Tech. in Computer Science, 2019

IISc Bangalore
B.E. in Computer Science, 2015

CEG, Anna University

Experience

Blue Scholar

IBM Research Labs

Aug 2019 – Dec 2020 Bangalore, India

Worked on various SOTA technologies like Data Readiness for AI, Federated Learning etc.

Member of Technical Staff - 1

Adobe Research Labs

Sep 2016 – Jul 2017 Bangalore, India

Worked on Adobe Muse, an application that generates responsive websites through a plug&play interface.

Software Development Engineer

Samsung Research and Development Institute

Aug 2015 – Aug 2016 Bangalore, India

Worked on Andorid UI development for Tuscany, a multi-function printer.

Updates

🔊 Presented a Tutorial on Gaussian Process in CS 337.
🔊 Delivered an invited talk on “Learning Recourse on instance environment” at Adobe Research
📃 Learning Recourse on instance environment to enhance prediction accuracy accepted at NeurIPS'22
📝 Serving in the PC of AAAI 2023 Research track
📝 Serving in the PC of AIMLSystems 2022 Research track
📣Teaching Linear Algebra for GATE in NPTEL. Videos available here!
📃 Federated Learning with Relevant Data paper accepted at AAAI'22
📝 KDD'21 tutorial on data quality
📝 Serving in the PC of CODS-COMAD 2021 Research track
📝 Serving in the PC of AIMLSystems 2021 Industry track
🏫 Started PhD with IIT Bombay
🔊 Delivered a guest lecture on “data quality for Machine Learning” at JDBI, Kolkatta
🔊 Delivered a guest lecture on “outsmarting outliers in social networks” at CEG on social networks
📃 One short paper accepted at ICDE 2021
📃 One research track paper accepted at AAAI 2021
🏫 Graduated with Masters from IISc and joining IBM Research
📃 One Reserach track paper accepted at WSDM 2020
📃 One Reserach track paper accepted at AAAI 2019

Projects

All2Vec

Identifying authors of an anonymized papers in a double blind conference setting.

Talks

Invited virtual talk at Adobe Research Bangalore.

A detailed talk on my Neurips'22 paper Learning Recourse on instance environment to enhance prediction accuracy.

Sep 23, 2022 12:00 AM Bangalore, India

Invited talk at the College of Engg. Guindy, Anna University.

A guest lecture delivered in the Machine Learning class on the topic - Radial Basis Netwoeks.

Mar 19, 2021 12:00 AM Chennai, India

KDD'20 Tutorial on Overview and importance of data quality for machine learning tasks

A tutorial on importance of data quality in ML. The talk surveys various techniques used to clean different enterprise data such as tabular and free flowing text.

Aug 23, 2020 12:00 AM Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining

Invited talk at JDBI, Kolkatta

An introductory talk on importance of data cleaning in Machine Learning applications.

Jul 1, 2020 12:00 AM Kolkatta, India

Invited talk at the College of Engg. Guindy, Anna University.

A guest lecture delivered in the Machine Learning class on the topic - Outsmarting the outliers in attributed network representation learning.

Feb 11, 2020 12:00 AM Chennai, India

Selected Publications

Lokesh Nagalapatti, Guntakanti Sai Koushik, Abir De, Sunita Sarawagi

November 2022 NeurIPS'22

Learning Recourse on Instance Environment to Enhance Prediction Accuracy

Learning recourse on the environment that generates data instances.

Lokesh Nagalapatti, Ruhi Sharma Mittal, Ramasuri Narayanam

February 2022 AAAI'22

Is your data relevant?: Dynamic selection of data in Federated Learning.

Using Reinforcement Learning based techniques to enable clients in Federated Learning derive updates only from relevant data.

Lokesh Nagalapatti, Ramasuri Narayanam

May 2021 AAAI'21

Game of Gradients - Mitigating irrelevant clients in Federated Learning

Using Cooperative Game theory approaches to mitigate the impace of noisy clients in Federated Learning.

Eitan Farchi, Ramasuri Narayanam, Lokesh Nagalapatti

April 2021 ICDE'21

Ranking Data Slices for ML Model Validation: A Shapley Value Approach

Shapley values based approach to detect slices of data that needs to be analyzed to debug the performance of a black-box ML model.

Sambaran Bandyopadhyay, Lokesh Nagalapatti, Saley Vishal Vivek (*), M N Murty

January 2020 WSDM'19

Outlier Resistant Unsupervised Deep Architectures for Attributed Network Embedding

Auto encoder inspired deep architectures for finding outlier robust network embedding for nodes in an attributed networks through adversarial training.

Sambaran Bandyopadhyay, Lokesh Nagalapatti, M N Murty

November 2018 AAAI'19

Outlier Aware Network Embedding for Attributed Networks

Finding outlier robust network embedding for nodes in an attributed network. We define outlierness for each node and learn them using closed form update rules.