Saving US from Opioid Addiction one Machine Learning solution at a time
This paper was written to record our experience as the 2nd place winners and recipients of a $15,000 reward in the 2019 Humana-Mays Healthcare Analytics Case Competition amongst 480 participating teams from across the US.
Language Modeling of Gutenberg Corpus
As a part of the final course project for Text Analytics, I decided to build language models using the freely available Gutenberg corpus. This blogs documents the experiments and findings of this project.
Text Classification of Amazon Reviews
I built four different models including Logistic Regression, SVM, fasttext and CNN using methodologies like TF-IDF, word2vec to classify Amazon Reviews as positive or negative.
ProScanner - Handwritten Text Recognition using Neural Networks
This blog documents my first deep learning project in which we developed a neural network model using CNNs and LSTMs for Optical Character Recognition to convert handwritten text images into digital text.
What's my house worth? - Building a full stack Flask app
This blog post documents my first foray into developing a full stack analytical pipeline to administer a machine learning solution including using AWS tools such as EC2, S3 and RDS for backend infrastructure and Flask for front end UI.
Luxury European Hotels: Predicting Guest Review Scores Based on Hotel Reviews
The goal of this project was to predict using statistical modeling and machine learning the score, on a scale from 1 (lowest) to 10 (highest) that a reviewer will give after a stay at a luxury hotel in one of six European cities.
Helping an Online bookstore quantify the impact of their promotion strategy
The goal of this project was to predict using statistical modeling whether a given customer will respond to a promotional email sent out by an online bookstore, and how much they will spend on buying books if they do respond.