Debanjan Datta

ddatta@vt.edu

@devDdata

408-314-0878

Linkedin

I am a PhD candidate in Discovery Analytics Center , Virginia Tech.

I am part of Discovery Analytics Center, and advised by Dr. Naren Ramakrishnan.

Prior to this I was a Software Engineer in Yahoo Inc.

Research Interests

Detecting Suspicious Timber Trades
- In Proceedings of the Thirty-Second Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-20), New York, NY, Feb 2020,
- Debanjan Datta, Naren Ramakrishnan and others.
ProtTox: Toxin identification from Protein Sequences
- In Proceedings of the 14th Machine Learning in Computational Biology (MLCB) Meeting, NeurIPS'19, Dec 2019
- Sathappan Muthiah, Debanjan Datta, Mohammad Raihanul Islam1, Patrick Butler, Andrew Warren, Naren Ramakrishnan [ Joint first authorship ]
Imputing Data for the Fragile Families Challenge: Identifying Similar Survey Questions with Semi-Automated Methods
- Socius, 2019
- Brian .J. Goode, Debanjan Datta and Naren Ramakrishnan

-----------------------

Role : Software Development Engineer (LEVEL 2)

Duration : January 2015 - July 2016

Team : Yahoo Sports, back-end/API

Test driven development and Agile methodology
Features delivered : Core sports web-services, Feed processing, API features (for enriched front end), Internal frameworks and tools, Database and API optimization (flexibility, scalability, performance)
Independently created a tool to ensure data coherence in ingestion pipeline handling terabytes of data with complete automation.
Worked on real time troubleshooting on live production issues

-----------------------

Role : SOFTWARE DEVELOPMENT ENGINEER (LEVEL 1)

Duration : July 2013 - December 2014

Team : Yahoo! Mobile Web, back-end/API

Test driven development and Agile methodology
Features delivered : Developed back-end platforms and applications (dynamic experience to users mobile phones across plat-
forms, devices,regions and languages.), API optimization, Internal frameworks & tools, CI/CD integration & migration
Worked on real time troubleshooting on live production issues.

-----------------------

Some of my projects :

NOAA Weather Ingestion
- https://github.com/ddatta-DAC/NOAA_Weather_Fetcher
- NOAA has an excellent source of weather data, across the globe , since 2001. This repository has the code to ingest data, and store it in mongodb after parsing the data.
Deep Auto Encoder Tutorial
- https://github.com/ddatta-DAC/ConvolutionalAutoencoder
- A simple tutorial I created to help anyone trying to implement a convolution autoencoder.
Time Series Analysis
- https://github.com/ddatta-DAC/Time-Series-Analysis
- This repository has around 8 deep learning models implemented to analyze their performance on predicting timeseries , using exogenous variables - akin to ARX (Auto regressive with exogenous).
- The data set used is stock market data. It essentially shows the performance of various models, and the code can be extended by anyone wishing to build upon it.