Data Scientist

• Mar 2022 - Present

Analyst

• Jul 2021 - Aug 2021

Student

• Jul 2017 - Jul 2021

Research Intern

• Apr 2020 - Jul 2020

I have worked under supervision of Dr. Asif Ekbal at AINLPML lab of IIT Patna. During training period I have developed a multilabel classification model for Kaggle Toxic Challenge Dataset. A simple interface has been developed for hosting the trained model which can be found here: https://abctoxic.herokuapp.com/ As a part of IITP AINLPML team I have taken part in development of attention layer based deep neural network model for 1st round of EmotionGIF Challenge (For SocialNLP 2020 held in conjunction with ACL 2020) and achieved 5 the rank in the leader board. For the second-round ensemble model was developed to improve performance of the model and we achieved 5th rank in the final leader board. Our system description paper can be found here: https://arxiv.org/abs/2012.12756 and also here:https://sites.google.com/view/emotiongif-2020/reports?authuser=0 As a part of my internship I briefly worked on data annotation where my responsibility was to manually assign the textual data according to their sentiment for further experiment. Show less

Deep Learning Research Intern

• May 2019 - Jul 2019

As a research intern I have investigated the possibilities of identification of native reader using deep learning algorithms on the basis of the data related to eye movement of the reader provided in GECO Corpus. Comparing performance of different deep learning models we concluded that bidirectional LSTM can outperform others in this task. As a research intern I have investigated the possibilities of identification of native reader using deep learning algorithms on the basis of the data related to eye movement of the reader provided in GECO Corpus. Comparing performance of different deep learning models we concluded that bidirectional LSTM can outperform others in this task.