Machine Learning Labs

A collection of hands-on Machine Learning lab notebooks covering core ML concepts, trained on real-world datasets. Implementations are done both from scratch and using scikit-learn. Completed as part of a 5th semester Machine Learning course.

---

📂 Contents

Notebook	Topics Covered
data_preprocessing.py	Handling missing values, encoding categorical features, data cleaning
Linear_Regression.ipynb	Simple & multiple linear regression from scratch, RMSE, R² score
Linear_Regression_Sklearn.ipynb	Linear regression using scikit-learn, feature scaling, evaluation
Logistic_Polynomial_Regression.ipynb	Logistic regression (scratch + sklearn), polynomial regression
KNN_Distance_Metrics.ipynb	K-Nearest Neighbors with different distance metrics (Euclidean, Manhattan)
KNN_SVM_Ensemble.ipynb	KNN vs SVM comparison, ensemble methods (Bagging, Boosting, XGBoost)
K_Means_Clustering.ipynb	K-Means clustering from scratch, elbow method, DBSCAN
KMeans_PCA.ipynb	Dimensionality reduction with PCA, visualizing clusters
Descision_Trees.ipynb	Decision Tree classifier with entropy criterion, confusion matrix, accuracy score & tree visualization

---

📊 Datasets Used

These labs use real-world datasets to train and evaluate models — not just toy data:

Dataset	Used For
1000 Companies	Multiple Linear Regression (profit prediction)
Car Price Prediction	Regression (predicting used car prices)
House / Admission Predict	Linear Regression (admission chance prediction)
Head & Brain	Simple Linear Regression (head size vs brain weight)
Medical Cost Personal	Regression (insurance cost prediction)
Mall Customers	K-Means Clustering (customer segmentation)
Telco Customer Churn	Classification (churn prediction)
Income Evaluation	SVM classification
Social Network Ads	Logistic Regression, KNN, SVM
Mushroom Dataset	KNN & Ensemble classification
COVID-19 Dataset	Data analysis & visualization
Iris	Clustering, PCA visualization
Bill Authentication	Descision Trees

---

Tech Stack

• Python 3
• NumPy — numerical computations
• Pandas — data manipulation
• Matplotlib / Seaborn — data visualization
• Scikit-learn — ML models and evaluation

---

Concepts Covered

• Supervised Learning — Linear Regression, Logistic Regression, Polynomial Regression, KNN, SVM, Descision Trees
• Unsupervised Learning — K-Means Clustering, DBSCAN
• Dimensionality Reduction — Principal Component Analysis (PCA)
• Ensemble Methods — Random Forests, AdaBoost, Gradient Boosting, XGBoost
• Model Evaluation — RMSE, R², Accuracy, Confusion Matrix, ROC Curve, K-Fold Cross Validation
• Data Preprocessing — Missing values, categorical encoding, feature scaling

---

How to Run

1. Clone the repository:

   git clone https://github.com/Zimal-Fatemah/Machine-Learning-Labs.git
   cd Machine-Learning-Labs

2. Install dependencies:

   pip install numpy pandas matplotlib seaborn scikit-learn

3. Open any notebook:

   jupyter notebook

Notebooks can also be opened directly in Google Colab.

---

📌 Notes

• Implementations are done both from scratch (using NumPy) and using scikit-learn, to build a solid understanding of the underlying math.
• Each notebook includes data loading, preprocessing, model training, and evaluation steps.

---

Author

Zimal Fatemah
BS Artificial Intelligence
GitHub