EDUCATION
Computer Science, Master of Science (M.S.)
University of Illinois at Chicago
August 2017 - December 2018 (Expected)
GPA: 3.9(4.0)
Relevant Coursework:
Graduate Student Representative for department of Computer Science
Electronics Engineering, Bachelor of Engineering (B.E.)
University of Mumbai
July 2013 - June 2017
Relevant Coursework:
Workshops: Cloud Computing, Indian Institute of
Technology, Delhi
Sir Ratan Tata Technical Scholarship for Engineering Students
EXPERIENCE
-
AUG 2018
Graduate Student Researcher
Artificial Intelligence and Robotics Laboratory
University of Illinois at Chicago
August 2018 - Present
Supervisor: Prof. Brian Ziebart
Tasks:
- BaxterTM Robot Motion Planning for Autonomous Execution of Self-Learned Tasks
- Master's Thesis Research Project to teach the Baxter robot to play level-1 games(e.g. pattern blocks)
-
JUNE 2018
Graduate Teaching Assistant
CS 480: Database Systems
University of Illinois at Chicago
Jun 2018 - August 2018
Department Supervisor: Prof. Robert Kenyon
Tasks:
- Coursework management
- Grading tests and assignments
-
MAY 2018
Graduate Student Researcher
Artificial Intelligence and Robotics Laboratory
University of Illinois at Chicago
May 2018 - June 2018
Mentor: Prof. Brian Ziebart
Tasks:
- Inverse Reinforcement Learning based Virtual Reality control interface for Baxter Robot
-
FEB 2018
Graduate Teaching Assistant
CS 422: User Interface Design and Development
University of Illinois at Chicago
Feb 2018 - May 2018
Department Supervisor: Prof. Robert Kenyon
Tasks:
- Coursework management
- Creating and managing tests
- Grading tests and assignments
- Classroom studio sessions
-
FEB 2016
Student Participant and Course Co-creator
Stanford Scholar
Stanford University
Feb 2016 – Dec 2017
Mentor: Dr. Rajan Vaish
Contributions:
- Admin for creation of the online course: “Data Science and Machine Learning using Python”.
- Co-created short talks on CS papers on Machine Learning, HCI, Crowdsourcing, Computer Vision, and Security.
-
MAY 2015
Student Developer and Researcher
Stanford CrowdResearch (Daemo Crowdsourcing Platform)
Stanford University
May 2015 – Dec 2017
Mentor: Prof. Michael Bernstein, Dr. Rajan Vaish
Contributions:
- Member of initial front end design and development team
- Contributed to platform pilot studies, developing the “Boomerang taskfeed mechanism”, and “Open-Gov model”
- Member of team responsible for designing and developing constitution model for Daemo
-
DEC 2014
Android Developer Intern
Wegilant, Mumbai
Dec 2014 – Jan 2015
Role:
- Android Developer Intern at Wegilant, a company that provided Security systems and infrastructure for organizations
SKILLS
SOFTWARE SKILLS
Python
Scikit Learn
TensorFlow
Keras
PyTorch
NLTK
C
C++
Java
HTML
CSS
JavaScript
Flask
Bootstrap
ROS
Android
MySQL
Git
TECHNICAL SKILLS
Machine Learning
Computer Vision
Data Mining
Natural Language Processing
Artificial Intelligence
Reinforcement Learning
Robotics
Crowdsourcing (HCI)
PROJECTS
Robot Motion Planning for Autonomous Execution of Learned Tasks
Teaching Baxter Humanoid Robot to play Level-1 games
Advisor: Prof. Brian Ziebart
Salient Features:
- Training Baxter Robot to learn Block Slot Sorter game
- Using approximate Q-Learning
- Grasp-detection by using end-effector to Cartesian distance mapping, with Computer Vision
Technologies used
- Python
- OpenCV
- Keras
- ROS
Keywords: Reinforcement Learning, Deep Learning, Computer Vision, Artificial Intelligence, Game Theory
Automatic Image Captioning
Automatic Image Captioning for COCO dataset (20 GB), using Inception V3 Model, and Deep LSTMs
Salient Features:
- InceptionV3 for feature Extraction (Keras)
- Deep LSTMs for caption generation
- 53% test accuracy, with just 1 hour of training
Results:
- Python
- Sci-kit Learn
- Keras
Keywords: Deep Learning, Machine Learning, Computer Vision, Text Generation
Aspect Based Sentiment Classification
Aspect Based Sentiment Classification of food and text reviews using Scikit Learn
Mentor: Prof. Bing Liu
Salient Features:
- Used 13 different classifier models
- Average accuracy achieved: 73%
- F-score: 0.83 (+ class), 0.77 (- class)
Technologies used
- Python
- Scikit Learn
- NLTK
Keywords: Text Mining, Sentiment Analysis, Opinion Mining
MonoRL: Reinforcement Learning Agent for Intelligent Monopoly
Reinforcement Learning agent for playing Monopoly, modelled as a Markov decision Process (MDP)
Mentor: Prof. Plamen Petrov
Salient Features:
- Implemented 3 agents:
- ɛ-Greedy Q(λ)-Learning Agent
- Fixed Policy Agent
- Random Agent
Results:
- Won 61/100 games
- Relative Assets: 43%
- Peak Relative Money: 47%
Keywords: Reinforcement Learning, Artificial Intelligence, Game Theory, Non-Deterministic Problems
CereBro: Intuitive Scheduling Tool for Direct Knowlege Sharing
Bridging the gap between students who seek knowledge, and students who possess knowledge
Salient Features:
- Connect knowledge seeking students with their mentor counterparts
- Modified matching algorithm to bring students together
- Schedules meetings automatically
Technologies used
- Android Platform
- Retrofit 2.1
- Ruby on Rails
Keywords: Human-Computer Interaction, Education, Mobile Application
Triton: Predictive Assistance for Amateur Stock Traders
7-day predictions of opening stock prices for select companies listed under the NYSE
Salient Features:
- Used 15 years stock data for 10 different companies
- 7-day predictions of opening stock prices
- Accuracy: 98% on test data
Technologies used
- Python
- Scikit Learn
- Matplotlib
Keywords: Data Mining, Predictive Analysis, Supervised Learning, Mathematical Modelling
Segregator Robot
Segregator Robot using Firebird V Robotic Research Platform
Categorizes and transports objects to assigned drop-off areas (similar to dock cranes)
Salient Features:
- Categorizes objects based on shape and color
- Transports objects to assigned drop-off areas, by navigating path grid
Technologies used
- C
- Atmega controller
- Matlab
Keywords: Robotics, Algorithms, Computer Vision