Probability distribution fitting, XGBoost and Backwards Elimination Regression for feature selection; Multiple and Simple Linear Regression for modeling; Principal Component Analysis (PCA) for dimensionality reduction, then K-Means for clustering; 3D matplotlib plotting; T-SQL for business analysis. 

For forecasting supply chain production levels: ARIMA model and Facebook Prophet. For 1975 labor force participation: logistic regression, XGBoost, and Keras Neural Network classification with Google's TensorFlow.

Statistical Analysis of A/B Test, Causal Inference, and Web Recommendation. Consumer engagement segmentation via K-Means Clustering.

For predicting insurance loss claim amount: XGBoost, support vector (polynomial), standard and Lasso regression. For predicting insurance loss: logistic regression and XGBoost. For news headlines: Multinominal Naive Bayes.

Data science presentation on what makes a playlist great/successful. Includes: user research survey, exploratory data analysis, and standard linear regression.

Analyzed tipping rates and other metrics for delivery company. Found that one of the wealthiest towns in America had the lowest tipping rate. Recommended A/B test to see if displaying a short story on the delivery person at checkout might lead to more tips.

SQL + Python data analysis of a news company's data from the Facebook API. Looked at the most-consumed media content, devised new metrics, and offered recommendations.

Brief script reading a csv into a pandas dataframe, gaining insights by querying via SQLite, then visualizing results with matplotlib.

Brief python script answering a polling exercise question. Statistical inferences: One sample proportion z interval, two sample proportion z test, two sample mean t test.

This was a fun Python script I wrote which found my longest periods consecutive weight loss or weight gain. (Once upon a time, I was 265 pounds, until Swine Flu helped me lose weight).