Locally Weighted Regression

Requirements

• Python version: 3.6.3

Usage

Run the script:

python Assingment1_Eda.py

This will:

• Print the Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Square Error (RMSE) values to the console.
• Generate a result.csv file containing all test instances, along with their index, feature value, actual value, and predicted value.

Hyperparameters

The following hyperparameter values were used for the locally weighted regression model:

Hyperparameter	Value
c (bandwidth)	0.2
Learning rate	0.1
Nearest neighbours	25

Output

• Console output: MAE, MSE, and RMSE values.
• result.csv: Test instance index, feature value, actual value, and predicted value for every test instance.

A sample result.csv (generated from a prior run) is included in this repository.

Notes

Additional details about the implementation are documented as comments within the code.