Shounak Mitra

Major research areas included machine learning, advanced analytics, computer vision, image processing. data wrangling and modeling, android app development, social network analysis. - Developed a predictive model that detects, tracks and estimates the number of people walking in a hallway by analyzing the footstep vibration signals, picked up by an array of vibration sensors. - Improved the average daily traffic estimate on Memorial Bridge in Portsmouth-NH, by reducing the misclassification among vehicles by 25% by developing an automated system for object detection, tracking and predictive classification using an existing network of traffic cameras installed by the Department of Transport. - Designed a crawler to read, analyze and classify tweet sentiments into decisions, using Neural Network, for HVAC monitoring. - Worked on developing mobile apps (android) for students to help them design and analyze virtual 2-D bridge trusses and visualize the stresses and strains in the compression and tension truss members corresponding to various loading combinations.

- Generated statistical models to design a damage probability matrix that assigns an alpha numeric value to the damage state of a structure and helps in predictive modeling of loss estimates due to an earthquake event. - Led a team of 6 to automate a seismic risk estimation system, originally done using MS-Excel, that reduced the processing time by 900% and saved over 20% of project budget allocated for risk estimation

Developed web based application projects for a NBFC. Was responsible for designing framework and architecture for the projects as well as detailed project estimation of time line as well as resources.