Data Scientist
New York, Ny, Us
The Metis Data Science Bootcamp is a selective, 12-week immersive program that provides instruction and training in all facets of data science and machine learning. A few of the projects I worked on are listed here:• Personalized Movie RecommenderBuilt a movie recommender system using the MovieLens 20 million ratings dataset based on a collaborative filtering technique. I developed a novel approach to recommend movies to groups of two users watching together. Tools used: pandas, NumPy, SciPy, SVD, sparse matrices• Principal Component Analysis and Topic Modeling of RecipesExamined recipe data from ~12,500 recipes representing 25 different cuisine types and conducted PCA and LDA topic modeling. I created insightful visuals that demonstrated how the different cuisine types were related. Tools used: Yummly API, Requests, pandas, NumPy, matplotlib, scikit-learn, nltk, PCA, LDA, K-means clustering• Classifying Poisonous and Edible MushroomsPerformed classification of a mushroom dataset containing 8000 total samples. Achieved perfect classification accuracy of 1.0 to predict poisonous and non-poisonous mushrooms. Tools used: pandas, NumPy, matplotlib, seaborn, scikit-learn, logistic regression, KNN, decision trees• Linear Regression to Predict House PricesScraped home listing data of 500 recently sold homes located in Pleasanton, CA from Zillow.com. Performed linear regression analysis to accurately predict with a test set R2 score of 0.87 the home sale price based on 7 different input features. Tools used: pandas, Beautiful Soup, Selenium, Regular Expressions, matplotlib, seaborn, scikit-learn