🤖 Machine Learning with Python

A collection of end-to-end Machine Learning classification projects built with Python and scikit-learn, covering two real-world healthcare and environmental datasets. Each notebook walks through the complete ML pipeline — from raw data exploration and preprocessing all the way to training, evaluating, and comparing 7 different classification models.

---

📂 Repository Structure

Machine_learning_with_python/
├── Diabetes_Preprocessing_7_models.ipynb        # Diabetes prediction — full ML pipeline
├── water_potability_prediction_7_models.ipynb   # Water potability prediction — full ML pipeline
└── water_potability.csv                         # Dataset for the water potability project

---

📓 Projects Overview

🩺 Project 1 — Diabetes Prediction

Notebook: Diabetes_Preprocessing_7_models.ipynb

Predicts whether a patient is diabetic based on diagnostic measurements. Uses the well-known Pima Indians Diabetes Dataset.

Dataset Features

Feature	Description
Pregnancies	Number of pregnancies
Glucose	Plasma glucose concentration (2-hour OGTT)
BloodPressure	Diastolic blood pressure (mm Hg)
SkinThickness	Triceps skin fold thickness (mm)
Insulin	2-hour serum insulin (mu U/ml)
BMI	Body mass index (weight/height²)
DiabetesPedigreeFunction	Diabetes pedigree function (genetic risk)
Age	Age in years
Outcome	Target — 1: Diabetic, 0: Non-diabetic

Pipeline Steps

• Exploratory Data Analysis (EDA) — distributions, correlations, class balance
• Handling biologically invalid zero values (e.g., zero Glucose, zero BMI) → replaced with NaN
• Missing value imputation (mean/median)
• Feature scaling — StandardScaler / MinMaxScaler
• Train/test split
• Training and evaluating 7 classification models
• Comparing model performance with metrics and visualizations

---

💧 Project 2 — Water Potability Prediction

Notebook: water_potability_prediction_7_models.ipynb
Dataset: water_potability.csv

Predicts whether water is safe to drink based on physicochemical measurements.

Dataset Features

Feature	Description
ph	pH level of water (0–14)
Hardness	Water hardness (mg/L)
Solids	Total dissolved solids (ppm)
Chloramines	Chloramines concentration (ppm)
Sulfate	Sulfate concentration (mg/L)
Conductivity	Electrical conductivity (μS/cm)
Organic_carbon	Organic carbon (ppm)
Trihalomethanes	Trihalomethanes concentration (μg/L)
Turbidity	Water turbidity (NTU)
Potability	Target — 1: Potable, 0: Not potable

Pipeline Steps

• EDA — feature distributions, missing values analysis, class imbalance
• Missing value imputation
• Feature scaling and normalization
• Train/test split
• Training and evaluating 7 classification models
• Model comparison with evaluation metrics

---

🧠 Models Used (Both Projects)

Both notebooks train and compare the following 7 classifiers using scikit-learn:

#	Model	Type
1	Logistic Regression	Linear
2	K-Nearest Neighbors (KNN)	Instance-based
3	Support Vector Machine (SVM)	Kernel-based
4	Decision Tree	Tree-based
5	Random Forest	Ensemble (Bagging)
6	Gradient Boosting	Ensemble (Boosting)
7	Naive Bayes	Probabilistic

---

📊 Evaluation Metrics

Each model is evaluated using:

• Accuracy — overall correct predictions
• Precision — how many predicted positives are actually positive
• Recall — how many actual positives were correctly identified
• F1-Score — harmonic mean of Precision and Recall
• Confusion Matrix — visualizes TP, TN, FP, FN
• ROC-AUC Curve — model discrimination ability

---

🛠️ Tech Stack

Library	Purpose
pandas	Data loading, manipulation, preprocessing
numpy	Numerical operations
matplotlib	Data visualization and plots
seaborn	Statistical visualizations
scikit-learn	ML models, preprocessing, evaluation metrics
Jupyter Notebook	Interactive development environment

---

⚙️ Setup & Installation

1. Clone the repository:

git clone https://github.com/Mortezamohasebati/Machine_learning_with_python.git
cd Machine_learning_with_python

2. Install required packages:

pip install pandas numpy matplotlib seaborn scikit-learn jupyter

3. Launch Jupyter Notebook:

jupyter notebook

4. Open either notebook and run all cells (Kernel → Restart & Run All)

---

🔄 ML Pipeline Overview

Both projects follow this standard pipeline:

Raw Data
   │
   ▼
Exploratory Data Analysis (EDA)
   │
   ▼
Data Preprocessing
 ├─ Handle missing / invalid values
 ├─ Feature scaling (StandardScaler / MinMaxScaler)
 └─ Train / Test split (typically 80/20)
   │
   ▼
Model Training (7 classifiers)
   │
   ▼
Evaluation & Comparison
 ├─ Accuracy, Precision, Recall, F1
 ├─ Confusion Matrix
 └─ ROC-AUC Curve
   │
   ▼
Best Model Selection

---

📚 Concepts Covered

• Binary classification
• Exploratory Data Analysis (EDA)
• Data cleaning — handling missing and invalid values
• Feature scaling — StandardScaler, MinMaxScaler
• Supervised learning — 7 classification algorithms
• Model evaluation — Accuracy, Precision, Recall, F1, ROC-AUC
• Confusion Matrix interpretation
• Comparing multiple models on the same dataset

---

📜 License

This project is open source and free to use for educational and research purposes.