Chetan Kumar

I am ambitious PhD student in Engineering and Applied Sciences (Computer and Information Science) at University of Massachusetts, US. My research foucses on Deep Learning, Computer Vision, Transfer Learning and Data Visualization. My Masters project was "Skeleton Based Action Recognition using Convolutional Neural Network" using Machine Learning under the supervision of Dr.Ming Shao.

Neural and Intelligent System Lab
University of Massachusetts


University of Massachusetts, Dartmouth US

PhD in Engineering and Applied Science
(Computer and Information Science) September 2018 - Present

University of Massachusetts, Dartmouth US

Masters in Data Science August 2018

Shaheed Zulfikar Ali Bhutto Institute of Science and Technology, Karachi Pakisan

Masters in Computer Science June 2016

Shaheed Zulfikar Ali Bhutto Institute of Science and Technology, Karachi Pakisan

Bachelors in Computer Science June 2014


University of Massachusetts, Dartmouth (Department of Nursing Sciences)

Data Analyst September 2017 - May 2018

Responsible for development of prediction model for successful NCLEX-RN candidates using Nursing Admissions Data and Courses Grade Sheet

Web Enthusiasts

Software and Web Developer June 2014 - October 2015

Responsible for Design, Development, Deployment and Maintenance of Software and Web Applications


Information Technology (IT) Department -- Internee June 2013 - August 2013

Worked as IT Assistant and Responsible for Troubleshooting of Computers, Printers and other hardware

Pakistan Telecommunication Limited

Pakistan Internet Exchange -- Internee June 2012 - August 2012

Responsible for Design and Development of Web Portal with Database to transfer their Spread Sheet Data over databases for their internal use


Programming Languages

Python R Matlab HTML/CSS/JS C(Parallel Programming) C++

Data Tools

Tableau Numpy Pandas D3 Rapid Miner MySQL

Machine Learning Packages

Keras Tensor Flow Matlab LibSVM

Other Tools

Git Latex Linux Windows


Skeleton Based Action Recognition using Convolutional Neural Network:

A skeleton based live working model for action recognition was developed to present the biasness present between synthetic and non-synthetic data. For this purpose Kinect V2 is used to get Skelton representation of 2 persons interactions. View

Data Visualization

Worked on H-1B Visa Petitions data set from Visualized data for number of visa petitions from all over the US, Number of applicants in each job category in each state, Number of incoming international students and number of visa petitions each year. View

Business Analytics and Data Mining

Worked on KKBox’s Churn Prediction Challenge from, to explore data using different visualization to find out which variables are significant in predicting churn. Used different machine learning approaches to make a model of churn prediction on the given data. Report and Code

Text Processing and Text Mining

Jupyter Notebook was configured on Stampede (Super Computer) to access it on local machine for performing the text processing and text mining on unstructured data using the Python NLTK library Code

Computational Reproducibility

Reproduced Paper “Real Time Robust L1 Tracker Using Accelerated Proximal Gradient Approach” using Matlab and Reproduced Papers “Incremental Learning for robust visual tracking” and “Real-time tracking via online boosting” for Comparing Experimental Results. Drawn relationship between Original Results and Reproduced Results by producing Charts and Table of the Experiments Performed. Report and Results


Conference Paper

D. Kumar, C. Kumar and M. Shao, "Cross-database mammographic image analysis through unsupervised domain adaptation" 2017 IEEE International Conference on Big Data (Big Data), Boston, MA, USA, 2017 Paper PPT Code


2017 NECV: Cross-Database Mammographic Image Analysis through Unsupervised Domain Adaptation PPT Poster


Conference Reviewer / External Reviewer

2017 - Association for Advancement of Artificial Intelligence (AAAI)

2018 - International Joint Conference on Artificial Intelligence (IJCAI)

2018 - Journal of Electronic Imaging (JEI)