python_AI_Scripts

This repository contains a collection of Python scripts covering data preprocessing, machine learning, regression, and algorithm visualization.
The scripts were created for coursework and experiments in AI and ML.

⚠️ Note: Some of these scripts rely on datasets (.csv / .xlsx) that are not included in this repository due to availability or licensing restrictions.
If you want to run those scripts, you will need to prepare or substitute similar datasets.

---

📂 Scripts Overview

1. data_cleaning.py

• Cleans and merges raw datasets (gas flows, process parameters, and specs).
• Handles outliers using the Interquartile Range (IQR).
• Imputes missing values with column means.
• Normalizes numerical features for machine learning.
• Saves a cleaned dataset as CSV.
  ⚠️ Requires gas_flows.xlsx, process_parameters.xlsx, and IOP_specs.xlsx, which are not provided:contentReference[oaicite:0]{index=0}.

---

2. regression_model_fitting.py

• Loads the cleaned dataset.
• Fits Random Forest Regressors to predict C and MET.
• Evaluates model performance using Root Mean Squared Error (RMSE).
  ⚠️ Requires cleaned_merged_data.csv, which is generated by data_cleaning.py:contentReference[oaicite:1]{index=1}.

---

3. manual_regression_model.py

• Implements Polynomial Regression (degree = 2) manually using matrix algebra (Normal Equation).
• Computes weights, predictions, and RMSE.
  ⚠️ Requires an external data.csv file:contentReference[oaicite:2]{index=2}.

---

4. polynomial_regression_plot.py

• Fits a Polynomial Regression model with scikit-learn.
• Splits dataset into train/test.
• Calculates RMSE, predicts values, and plots fitted curves.
  ⚠️ Depends on data.csv:contentReference[oaicite:3]{index=3}.

---

5. polynomial_regression_fullfit.py

• Fits polynomial regression on the entire dataset.
• Evaluates RMSE across all points.
• Plots regression curve.
  ⚠️ Requires data.csv:contentReference[oaicite:4]{index=4}.

---

6. knn_classifier.py

• Custom implementation of K-Nearest Neighbors (KNN) from scratch.
• Includes performance metrics (confusion matrix, accuracy, precision, recall, F1).
• Uses a hardcoded sample dataset, so it can run without external data:contentReference[oaicite:5]{index=5}.

---

7. ml_classifiers_comparison.py

• Compares multiple classifiers on the Iris dataset:
  • KNN
  • Decision Trees
  • Random Forests
  • Naive Bayes
  • Linear SVM
  • Kernel SVM
• Reports classification accuracy for each.
  ✅ Uses scikit-learn’s built-in Iris dataset, no extra files needed:contentReference[oaicite:6]{index=6}.

---

8. ml_hyperparameter_tuning.py

• Performs hyperparameter tuning for KNN, Decision Trees, Random Forests, and SVMs.
• Finds optimal values for k, tree depth, estimator count, kernel, and gamma.
  ⚠️ Requires data.xlsx containing moon, circle, and blob datasets:contentReference[oaicite:7]{index=7}.

---

9. knights_tour_gui.py

• Tkinter GUI to visualize the Knight’s Tour problem on chessboards (5x5 → 8x8).
• Allows setting starting position and animation speed.
• Animates knight’s path with board highlighting.
  ✅ Runs without datasets.
  ⚠️ Requires a local knight.png image to display the knight piece:contentReference[oaicite:8]{index=8}.

---

🚀 How to Run

Clone the repository and run any script:

git clone https://github.com/yourusername/python_AI_Scripts.git
cd python_AI_Scripts
python knn_classifier.py