MACHINE LEARNING DIABETES PREDICTION 📊

Introduction 📝

This project focuses on predicting diabetes using machine learning techniques. The dataset includes various health and demographic attributes, which are used to predict whether an individual is diabetic.

Task 🎯

The primary task is to build a predictive model that determines if an individual has diabetes based on features such as age, sex, cholesterol level, BMI, smoking habits, and more.

Dataset 📁

The dataset contains the following columns:

• Age: Age of the individual.
• Sex: Gender of the individual.
• HighChol: Indicator if the individual has high cholesterol.
• CholCheck: Indicator if the individual had a cholesterol check in the past five years.
• BMI: Body Mass Index.
• Smoker: Indicator if the individual is a smoker.
• HeartDiseaseorAttack: Indicator if the individual has had a heart disease or attack.
• PhysActivity: Indicator if the individual engages in physical activity.
• Fruits: Indicator if the individual consumes fruits regularly.
• Veggies: Indicator if the individual consumes vegetables regularly.
• HvyAlcoholConsump: Indicator if the individual is a heavy alcohol consumer.
• GenHlth: General health indicator.
• MentHlth: Mental health indicator.
• PhysHlth: Physical health indicator.
• DiffWalk: Indicator if the individual has difficulty walking.
• Stroke: Indicator if the individual has had a stroke.
• HighBP: Indicator if the individual has high blood pressure.
• Diabetes: The target variable indicating if the individual has diabetes.

Models Used 🧠

In this project, four different models were implemented:

1. Logistic Regression
2. K-Nearest Neighbors (KNN)
3. Support Vector Machine (SVM)
4. Neural Network (NN)

Exploratory Data Analysis (EDA) 📊

EDA was conducted to understand the distribution of the features and identify any correlations or patterns. Visualizations were created using Matplotlib and Seaborn.

Performance Metrics 📈

The following metrics were used to evaluate the performance of the models:

• Accuracy Score
• F1 Score
• Confusion Matrix

Conclusion ✅

The results indicate that the models achieved different levels of accuracy and F1 scores. However, there is room for improvement, particularly in the fine-tuning of the model's hyper-parameters.

Libraries Used 📚

• pandas
• numpy
• matplotlib
• seaborn
• scikit-learn
• pickle

Future Work

Potential improvements could include experimenting with different models such as Random Forests or Gradient Boosting, and applying techniques like cross-validation and hyperparameter optimization.

Credits

The dataset used in this project was obtained from Kaggle.