jbbjr Goto Github PK

big-data

capi

CAPi (Climate Adjusted Pace individualized) is a predictive model trained on a users Strava data. It leverages optimal weather conditions from my own research to return a Climate Adjusted Pace (CAP) for an athlete. Similar to Gradient Adjusted Pace (GAP), this will help athletes train more efficiently in suboptimal weather conditions.

marathon-weather-analysis-and-performance-prediction

In this project, I utilize BigQuery, Selenium, Pandas, the Visual Crossing weather API, and Econometric techniques to determine the effects of weather on a runner's marathon performance.

nba-data-playbook

repository for B775 project @ Boston University MSBA

save-the-frogs

The aim of the project is to classify risk levels of species based on economic and environmental factors. The dataset is constructed from multiple APIs such as EDGAR, the IUCN Red List and the World Bank. In here we utilize sklearn fundamentals such as pipelines, scoring metrics, search methods for hyper parameter tuning, modeling, and ensembles.

sea-temp-lstm

project repository

sound-vs-engagement

This project focuses on conducting an experiment, collecting data, and analyzing its results. A key focus is randomization, to ensure heterogenous treatment effects. Randomization was conducted simply via shuffling, and tested via a proportions z-test. Evaluation methodologies included difference in means and OLS with fixed effects

spotify-eda-and-clustering

This project focuses on Exploratory Data Analysis (EDA) in Jupyter Notebook. Here, we utilize ETL, fundamental pandas knowledge for data science, as well as some simple unsupervised machine learning in the form of K-means clustering. The scope of this project is to try and derive value from the metric 'Track Popularity' in the Spotify API.

yelp

This project focuses on general unsupervised machine learning techniques, such as clustering via K-Means and dimensionality reduction via PCA and TSNE. It also includes fundamental NLP techniques within the scope of the nltk library. The scope of the project is to derive some kind of business value from the Yelp Academic Dataset via these methods.