🩺 Diabetes Prediction using Support Vector Machine (SVM)

A machine learning project that predicts whether a person is diabetic or not based on diagnostic health measurements, using the PIMA Indians Diabetes Dataset and a Support Vector Machine (SVM) classifier.

---

📌 Project Overview

This project aims to build a binary classification model that can predict the likelihood of diabetes in a patient. It covers the full ML pipeline — from data loading and exploration to model training, evaluation, and a real-time predictive system.

---

📂 Dataset

• Name: PIMA Indians Diabetes Dataset
• Source: Originally from the National Institute of Diabetes and Digestive and Kidney Diseases
• Records: 768 patients (female, at least 21 years old of Pima Indian heritage)
• Target Column: Outcome — 0 = Non-Diabetic, 1 = Diabetic

Features

Feature	Description
Pregnancies	Number of times pregnant
Glucose	Plasma glucose concentration
BloodPressure	Diastolic blood pressure (mm Hg)
SkinThickness	Triceps skinfold thickness (mm)
Insulin	2-Hour serum insulin (mu U/ml)
BMI	Body mass index
DiabetesPedigreeFunction	Diabetes pedigree function score
Age	Age in years

---

🛠️ Tech Stack

• Language: Python 3
• Libraries:
  • numpy — numerical operations
  • pandas — data loading and analysis
  • scikit-learn — preprocessing, model training, and evaluation

---

🔄 Project Workflow

1. Data Collection & Loading
        ↓
2. Exploratory Data Analysis (EDA)
        ↓
3. Data Standardization (StandardScaler)
        ↓
4. Train-Test Split (80% train / 20% test)
        ↓
5. Model Training (SVM — Linear Kernel)
        ↓
6. Model Evaluation (Accuracy Score)
        ↓
7. Predictive System (single-instance prediction)

---

📊 Model Details

• Algorithm: Support Vector Machine (SVM) with a linear kernel
• Train-Test Split: 80% training, 20% testing (stratified split, random_state=2)
• Preprocessing: Feature standardization using StandardScaler

---

✅ Results

Dataset	Accuracy
Training Data	~78–79%
Test Data	~77–78%

Exact values may vary slightly based on environment.

---

💡 Sample Prediction

input_data = (5, 166, 72, 19, 175, 25.8, 0.587, 51)
# Output: The person is diabetic

The input is reshaped, standardized using the same scaler fitted on training data, and passed to the trained SVM model for prediction.

---

🚀 How to Run

1. Clone the repository

   git clone https://github.com/Mdwali79986/PROJECT_Diabetes-Prediction-using-Machine-Learning.git
   cd PROJECT_Diabetes-Prediction-using-Machine-Learning

2. Install dependencies

   pip install numpy pandas scikit-learn

3. Add the dataset

   • Download diabetes.csv from Kaggle – PIMA Diabetes Dataset
   • Place it in the project directory

4. Run the notebook

   jupyter notebook Project_Diabetes_Prediction.ipynb

---

📁 Project Structure

diabetes-prediction/
│
├── Project_Diabetes_Prediction.ipynb   # Main notebook
├── diabetes.csv                        # Dataset (not included — add manually)
└── README.md                           # Project documentation

---

👤 Author

Md Waliullah

• 📧 mdwali79986@gmail.com
• 🔗 LinkedIn
• 💻 GitHub

---

📃 License

This project is open-source and available under the MIT License.