DATA SCIENCE PROJECTS
THE UNITED STATES OF COVID
LIVE DASHBOARD TRACKING COVID-19 SPREAD ACROSS THE USA AT COUNTY LEVEL
COVID-19 brought the world to a screeching halt. It is the invisible enemy surrounding us.
To help visualize it, I created a live website with an interactive dashboard that automatically tracks daily COVID-19 case/death numbers as published by the New York Times.
Tools Used: Python, Plotly-Dash, Heroku, Plotly, Folum, Pandas, Git
PUNCH BUDDY BOXING SYSTEM
wearable Closed Loop Feedback Training System Implementing a Tensor Flow LSTM Recursive Neural Network
How do the top performing athletes get to the top of their game and consistently improve their performance?
They look at the data.
Quantification of workouts and training allows the athlete to identify strengths, weaknesses, and improvements of their chosen craft.
Tools Used: Python, IMU Sensors, Tensor Flow Neural Network, Seaborn, Pandas, Git
THE GULAG GUNSMITH
CALL OF DUTY MODERN WARFARE
Uses Computer Vision to determine the +/- of each stat
I've been playing first person shooters for a long time (shoutout to the GOAT CS 1.6). Gaming weapon customization has come a long way from then.
In the latest Call of Duty : Modern Warfare they take it to the next level.
Each weapon can have up to 5 attachments on it to tweak the stats of the weapon to better match either your playstyle or to make a special purpose build for a certain scenario.
Tools Used: Python, OpenCV, Plotly, Pandas, Git
WHAT WE TALKIN’ BOUT
NLP (Natural Language Processing) Analysis of Lyrical Data of Regional Rappers Across America
I wanted to look into the commonalities / differences of lyrics produced by songs of rappers from different regions across America to see if the uniqueness of each region can be quantified.
Historically there has always been the West Coast vs East Coast rap beef but in the recent years there has been the boom of artists from the Mid West as well as Down South.
Tools Used: Python, AWS RDS, Beautiful Soup, PostgreSQL, NLTK, TF-IDF, Cosine-Similarity, Seaborn
UFC FIGHT VICTOR PREDICTOR
UFC Matchup Winner Made From a Statistically Trained Gradient Boosted Classification Model
With available historical UFC fight data from (www.UFCstats.com), this information can be used to train a statistical model to predict (with hopefully some degree of certainty) the winner of an upcoming fight.
Tools Used: Python, Random Forest Regressor, KNN, Pandas, Seaborn