Disaster Response Classification - Text messages as input to a Multi-Label classification model to identify the diasaster category. A ETL and ML pipeline to process class imbalanced data from Appen and improved f1-score of model using undersampling.
Starbucks Promotional Analysis - Achieve an optimal promotional strategy by training a model based on the promotional data and hypothesis testing.
NYC Salary Growth Predicator - Model to predict salary based various features and forcast expected growth of a job to help inform NYC budget. Exploratory Data Analysis of the data set from OpenData.
NYC Airbnb - Identify the most valuable property assets and strategies to maximize profits for property owners in NYC from a NYC airbnb Dataset.
Steam Charts - Model and forecast univariate time series data of player activity in various video games with Autogression, LSTM and FB Prophet.