Projects
These are some of my projects I have completed!
Predicting if a Student will Graduate or Dropout
One of the projects I have worked on involved predicting student dropouts and academic success, in Python. I am able to predict if a college student is at risk of dropping out. I used several skills I learned, such as doing descriptive analysis to find features that may be important. I also used a correlation matrix, standard deviation, and frequency proportion to look at the data. Then, I moved on to the data mining aspect. My partner and I used logistic regression, KNN, decision tree, random forest, and boosting tree models. We extracted the accuracy, recall, precision, and F1 scores from these models to determine the best one. We found that age of enrollment, having debt, and curriculum grades were the most important factors if someone graduates or not.
Finding relationship between gas prices and election season.
This project was done during election time to decide if “gas prices” were an important factor during debates and deciding who to vote for. To complete this project, I used R-Studio to code and used gas price information from the U.S Energy Information Administration. This data included averages for all gas throughout the U.S. I used regression techniques and data visualization to find that there is not a correlation. I found that the prices between winter gas and summer gas (different evaporation temperatures) was the cause for the variance in gas prices.