Abnormality Detection in Chest X-rays (CXR)

The goal of this project is passing a raw CXR image to our trained machine learning model(s) and get predicted abnormality detection bounding boxes overlaid on top of it as output.
Keywords: Computer Vision
Website   GitHub Repo

Face Mask Detector

The goal of this project is to identify whether a person is correctly wearing a mask or not. A person correctly wears a mask when the mask completely covers their mouth and nose.
Keywords: Computer Vision
Website   GitHub Repo

Fatality Prediction for Car Crash

In this project we are predicting fatality of person in car crash. The dataset used was highly imbalanced for 1:minority there were 46:majority classes.
Keywords: Machine Learning
Website   GitHub Repo

Capital Bikeshare

Capital Bikeshare is very popular and always in demand service. This project predicts the number of bikes to use per hour and have contingencies to fulfill the demand. Various factors such as weather, temperature, working, or non-working hour, the hour of the day are factors to determine the number of bikes.
Keywords: Machine Learning
Website   GitHub Repo

Forecasting Traffic Volume

In this project we are predicting traffic volume per day for the I94 interstate. To make accurate forecasts, 6 time series models were used and their performances were compared.
Keywords: Time Series Analysis
Website   GitHub Repo

Red Blood Cell Classification

The goal of this project is to identify whether a red blood cell is healthy or infected with any of the following stages of malaria: ring, schizont and trophozoite using a Multi-Layer Perceptron (MLP).
Keywords: Computer Vision
Website   GitHub Repo

American Time Use Survey

The purpose of American Time Use Survey (ATUS) is to develop nationally representative estimates of how people spend their time. The inspiration behind this project is to understand how daily activities of Americans differ by employment status, income, gender, etc.
Keywords: Data Mining
Website   GitHub Repo

Recipe Extractor

The goal of this project is to create a recipe knowledge database. This is achieved by identifying and extracting - actions to perform on ingredients, utensils used, and identify ingredients and quantities from food recipes.
Keywords: Natural Language Processing
Website   GitHub Repo

Entity Extraction from Reuters Dataset

Using spaCy extracted persons, location and organizations using Reuters dataset.
Keywords: Natural Language Processing
Website   GitHub Repo

Ironman Triathlon Competition Analysis

An Ironman Triathlon consists of a 2.4-mile swim, a 112-mile bicycle ride and a marathon run, raced in that order. An exploratory data analysis was performed on the data and key insights are summarized.
Keywords: Data Mining
Website   GitHub Repo