Student Of Data Science
Remote
Learned Python, SQL, applied statistics, and machine learning. Developed skills in data collection, wrangling, and visualization. Coursework included structured classes and assignments as well as self directed projects with advise and oversight from an industry mentor.Projects:Predicting the Price of Used Cars-Used Google Cloud Platform, Dask JupyterHub and SSH tunneling to reduce training time from > 12 hours on a local machine to < 30 minutes, allowing faster comparison and tuning of models during model selection.-Used a dataset of over 3 million rows to create a price prediction tool with a mean absolute percentage error of 9.1% using only a few key features.Clustering FIFA Players-Deployed machine learning techniques including K-Means and PCA, to identify meaningful clusters of players.-Analyzed and labeled the clusters using multivariate regression analysis, to turn model output into actionable insights for hiring managers.-Presented the results of the analysis using insightful visualizations a panel of stakeholders.Predicting Bike Share Equipment Inactivity-Tested a variety of classification models including Random Forest, SVM, Logistic Regression and XGBoost.-Aggregated a dataset of over 3 million rows deploying feature engineering techniques to reduce the dimensions of the data while retaining key information, lowering training time.-Utilized open data source and automated collection of regularly released reports.