Menu
A Tech Enthusiast | Ramping-up my Data-Related Skills

Disha Punjabi

Learn More About Me

EDUCATION

Masters In Science

COMPUTER SCIENCE

University Of Southern California, LA, USA

Bachelor of Technolgy

COMPUTER SCIENCE

Veermata Jijabai Tehnological Institute, Mumbai, India

EXPERIENCE

Big Data And Advertising Platform Intern, Roku

Roku, Los Gatos, California

APPLICATION DEVELOPER

CREDIT SUISSE

Investment Banking, Mumbai, India

I am a Grad student at the Univerity of Southern California

Actively Seeking Full-Time Opportunities starting May 2017

Interested in the Areas of General Software Development

Exploring the Machine Learning and Data Science Domain at USC

Having gained Relevant Experience in Both the Fields,
Looking for an interesting opportunity to grow and apply these skills






Click on the Project Title to know more!

Projects

Following is the summary of a few of my past/current work

Look-ALike Modeling

Built Look Alike models for Roku’s advertisement platform using supervised learning models, with a Lift score = 80%. Developed the data pipeline by dichotomization, normalization techniques as a part of feature engineering and enrichment. Created a demographics prediction model along with feature-engineering, exploratory analysis and custom-evaluation metrics

Web Mining/ Scraping

Mined the Dark Web to fetch,parse,process and visualize Weapons Data across the US to mine patterns indicating unlawful/ underage/ illegal sales of weapons.

Trade Bull

Used Supervised and Unsupervised learning techniques to perform Portfolio Hedging for cross-asset financial instruments. Used K-Means Clustering and Sentiment Analysis on Twitter Feed data to accomplish the same.

DCap

A scalable trading platform that provides users a chance to download and extend a generic C++ and IB based API to trade in Security Markets around the world.

Improvising Image Search On Large Databases

Improvised Image Based Search Techniques on Large Scale Databases by merging two Computer Vision Techniques - COntent Based Image Retrieval and Scale Invariant Feature Transform. A research paper describing the aproach was published in the IJEAT journal

StoryBoard

A real-time dashboard comprising of Visualization of
a. Aggregated Statistics of the underlying Market and Internal Trade Data
b. Instant Messaging Order Placement Platform for Clients
c. Real-time Alerting/ Notification based system

Leadership Skills:

  • Senator for Computer Science Department, Viterbi Graduate Students Association

  • Conducted Seminars for Junior Students in Undergrad School on C and C++

  • Coached Higher Secondary Students for the subjects Mathematics nd Chemistry for Engineering Entrance Exam

  • Mentored new Incoming Students at USC as a part of Spring 2016 Mentorship Program

  • Interests

  • In Leisure time, I enjoy playing the Guitar

  • I like to Cook (mainly Indian and Indo-Chinese cuisines)

  • My Creative Skills Include: Henna Art, Crosio Art

  • Other Interests- Reading (fiction & non-fiction), Dancing (Learning Salsa-level 1)