Metis is a full‐time 12 week ACCET‐accredited immersive data science program which focuses on programming (Python), project design, data acquisition and cleaning, machine learning, statistical modeling, and communication. 𝗟𝗲𝘁'𝘀 𝗕𝗲𝗮𝘁 𝗗𝗿𝗮𝗳𝘁𝗞𝗶𝗻𝗴𝘀: 𝗡𝗕𝗔 𝗗𝗮𝗶𝗹𝘆 𝗙𝗮𝗻𝘁𝗮𝘀𝘆 𝗣𝗼𝗶𝗻𝘁𝘀 𝗣𝗿𝗲𝗱𝗶𝗰𝘁𝗼𝗿Created a linear regression model to predict nightly DFS fantasy point production.• Scraped stats.nba.com using a combination of BeautifulSoup/Selenium and managed data throuhg a PostreSQL database. • Engineered lagged features on both individual and team aggregation levels for player statistics.• Created a Tableau dashboard to visualize key performance metrics for players. 𝗜𝘀 𝗶𝘁 𝗮 𝗕𝗮𝗻𝗴𝗲𝗿?Developed a supervised classification model using scikit-learn to predict if a song would be a hit.• Extracted the audio features and genres from 5,400 tracks released in 2018 through the Spotify Web API.• Generated positive class label by scraping Billboard Hot 100 weekly charts and fuzzy merging with Spotify data.• Optimized Random Forest classifier model for f-beta (β = 2) using RandomizedSearchCV to tune hyperparameters. 𝗣𝗮𝗿𝘀𝗶𝗻𝗴 𝗿/𝗻𝗯𝗮Applied unsupervised natural language processing techniques to parse out trending topics on the NBA subreddit. • Obtained 8.8 million subreddit comments using the Pushshift.io API.• Preprocessed, lemmatized, and applied stop words to corpus using NLTK , SpaCy, and scikit-learn's CountVectorizer.• Performed topic modeling through a combination of non-negative matrix factorization and CorEx anchoring to identify 6 distinct topics for marketing use.