Diabetes Prediction

This project aims to predict whether a person is diabetic or not based on several health-related features using machine learning techniques.

Dataset

The dataset used for this project is stored in the file diabetes.csv. It contains information about various health measurements for a group of individuals, along with an outcome label indicating whether they are diabetic or not.

Getting Started

To run this project, follow the steps below:

1. Clone the repository: git clone https://github.com/namanshah22/Diabetes-Prediction-ML
2. Install the required dependencies: pip install pandas numpy sklearn
3. Run the project

Preprocessing

The dataset is preprocessed before training the machine learning model. The following steps are performed:

• Load the dataset using pandas.
• Split the dataset into features (x) and the target variable (y).
• Standardize the features using StandardScaler from sklearn.preprocessing.

Model Training

A support vector machine (SVM) model with a linear kernel is used for training the classifier. The following steps are performed:

• Split the preprocessed data into training and testing sets using train_test_split from sklearn.model_selection.
• Initialize the SVM classifier with a linear kernel.
• Fit the classifier to the training data using the fit method.
• Predict the labels for the training data and calculate the accuracy score using accuracy_score from sklearn.metrics.

Model Evaluation

The trained SVM model is evaluated on the testing data to assess its performance. The following steps are performed:

• Predict the labels for the testing data using the trained classifier.
• Calculate the accuracy score of the model on the testing data using accuracy_score.

Prediction

You can make predictions on new data by providing the input sample. The following steps are performed:

• Prepare an input sample as a Python list or numpy array.
• Reshape the input sample to match the shape expected by the scaler and classifier.
• Transform the input sample using the pre-fitted scaler.
• Predict the label for the transformed sample using the trained classifier.
• Print the prediction result based on the predicted label.

Conclusion

This project demonstrates the use of a support vector machine model for diabetes prediction based on health-related features. The trained model achieves a certain accuracy on the testing data. Further improvements and optimizations can be explored to enhance the accuracy and generalizability of the model.