Political Candidate Education Prediction

Overview

This project predicts the education level of political candidates using a Random Forest Classifier. It analyzes various candidate attributes like criminal record, total assets, party affiliation, and state to make predictions.

Prerequisites

Ensure you have the following installed:

• Python 3.x
• Required Python libraries: pandas, matplotlib, seaborn, scikit-learn

File Description

• train.csv: Training dataset containing information about political candidates.
• test.csv: Test dataset for predicting education levels.
• my_submission_rf_improved_2.csv: CSV file with predicted education levels for the test dataset.

Code Explanation

1. Data Preprocessing:

   • Load and clean the training and test datasets.
   • Convert 'Total Assets' values to numeric format for analysis.
   • Perform basic data exploration and visualization to understand the data.

2. Feature Engineering:

   • Create new features based on candidate names and constituency preferences to improve model performance.

3. Data Encoding:

   • Encode categorical variables into numeric format using LabelEncoder for model compatibility.

4. Model Training:

   • Split the training data into training and validation sets for model evaluation.
   • Train a Random Forest Classifier on the training data to predict education levels.

5. Prediction:

   • Use the trained model to make predictions on the test set.
   • Convert numeric predictions back to the original education levels for interpretation.

6. Output:

   • Save the predictions to a CSV file for further analysis and submission.

Instructions

1. Clone the Repository git clone <repository_url>
2. Install Dependencies: pip install pandas matplotlib seaborn scikit-learn
3. Run the Code: python3 main.py
4. View Results:

• After running the code, check the generated my_submission_rf_improved_2.csv file for the predicted education levels.

Additional Notes

• Ensure that the training and test datasets are in the same directory as the script.
• Adjust hyperparameters of the Random Forest Classifier or try different machine learning models for experimentation.
• Explore additional feature engineering techniques for potentially improving model performance.