Linear Regression
- This refers to the Linear_Regression_Ecommerce file. Here, we look at data from an ecommerce website on indivisuals on their website or other platforms.
- It starts with some exploratory data analysis using matplotlib.pyplot and seaborn
- Then, we create a linear regression model to predict the yearly amount of time spent interacting with the company based on features such as the Time on the App or length of membership
- The Linear Regression Model is then evaluated using mean absolute error, mean square error, and root mean sqaure error. The root mean square error was looked at more heavily, because it punishes outliers like the mse, but is in units that are understanable
Logistic Regression
- This refers to the Logistic Regression ml model as per colab book Titanic.ipynb
- We then do some exploratory data analysis to gain general insight into the data
- Then, we clean up the data by dropping cols that have too many missing values, and imputing values for another column
- When creating the ML model, I had to deal with categorical features and multicolinearity
KNN
- This refers to the KNN_Project.ipynb file
- In this project, I looked at anonymous data and applied the KNN algorithm to it.
- I also iterated the algorithm multiple times with different k values to then find the k value that gives the lowest error rate.
Decision Tree's and Random Forests