Research Exprience

QAML (Question-writing Aided by Machine Learning)

Professor Jordan Boyd-Graber, University of Maryland, College Park

• Developed QAML (Question-writing Aided by Machine Learning), an interface for adversarial question-writing enhanced by the QANTA QA dataset. Trained and integrated language models in the interface for assisting users to write adversarial questions. Created modules for improving representation in the QA datasets.
  https://trick.umiacs.umd.edu/
  Keywords: BERT, DistilBert, Question-Answering, HCI, Web-Development
  
  

Bitcoin Data Analytics

Professor Ashutosh Bhatia, SDN Lab, BITS Pilani
Time: April 2020 - Nov 2020

• Created a spark-based framework for supporting large-size Bitcoin transaction graph data and did link prediction and graph embedding generation using Graph Convolution Neural Networks for suspect-community detection and transaction analysis
  Publication: “Bitcoin Data Analytics: Scalable techniques for transaction clustering and embedding generation,” 2021 International Conference on COMmunication Systems & NETworkS (COMSNETS), 2021, pp. 1-6, doi: 10.1109/COMSNETS51098.2021.9352922
  Keywords: Big Data, Spark, GraphNN, BlockChain, Clustering
  

Crop Recommender System

Professor Lavika Goel, Birla Institute of Technology and Science, Pilani
Time: Jan 2020 - Nov 2020

• Develop an intelligent Crop Recommendation System that suggests relevant crops for a given soil using satellite images by combining the expertise of Remote Sensing and Machine Learning algorithms.
  Publication: “Integrated Models of Machine Learning for Design of a Crop Recommendation System for Rajasthan, India” in Journal of the Indian Society of Remote Sensing (JISRS) (2021), (Under Review)
  Keywords: Remote Sensing, Machine Learning, CNN
  

Federated Learning

Professor Navneet Goyal, ADAPT Lab, BITS Pilani
Time: Aug 2020 - Dec 2020

• Developed an architecture to train a Lung Cancer classifier model in a Federated Learning setup. Applied Explainable AI to compare a model trained using Federated Learning vs conventional approach.
  Keywords: Federated Learning, Explainable AI, Computer Vision, CNN