This repository is dedicated to the class projects I have done and in the process of doing. These projects conists of simple data exploration, linear regression, logistic regressions, xgboost, roc curves, decsion trees, and importance plots. More information about these projects is listed in the html file I attached in the specific files.
Project 1: Is a simple data exploration as seen from the html file and consists of finding relationships between variables as well as fitting the model in a linear regression, elastic net, LASSO and decision tree. This data is a subset from the United Nations Department of Economics and Social Affairs world population prospect data. The original dataset contained approximately 43,473 observations, 55 indicators, and location notes. By specifically focusing on the Japanese population from 1950 to 2022, I was able to subset the data to a 73 55 dimension. Factors studied in this set include various quantitative variables, such as total population, population density, median age, etc.
- Project Part 1: Data Exploration
- Project Part 2: Linear Regression without train/test set
- Project Part 3: Various model application for variable selection and accuracy improvement
Project 2: This project consists of an xgboost to predict the highest level of injury you can obtain from a crash. The data was collected from the department of transportation from San Jose. Unfortunately, no imputations were done on this dataset.
Languages used: R with Rmd