DIABETES PREDICTION MODEL

Dataset to be used : https://www.kaggle.com/uciml/pima-indians-diabetes-database

1. Exploratory Data Analysis and Data Visualization

• General View
• Categorical Variables Analysis
• Numerical Variables Analysis
• Target Analysis

2. Data Preprocessing and Feature Engineering

• General View - Recap - Remember Dataset
• Outlier Analysis
• Missing Values Analysis
• Feature Creation
• Label and/or One Hot Encoding
• Standardization
• Save the Final Dataset --> Pickle Dataset

3. Modeling

• Logistic Regression
• Naive Bayes Classifier
• K-Nearest Neighbors Classifier
• Support Vector Machines
• Artificial Neural Network Models
• DecisionTreeClassifier
• BaggingClassifier
• RandomForestClassifier
• AdaBoostClassifier
• Gradient Boosting Classifier
• XGBoost - XGBClassifier
• LightGBM - LGBMClassifier
• CatBoost - CatBoostClassifier
• NGBoost - NGBClassifier

4. Pickle the Models, Saving the Model for later Use

5. Comparison of Metrics of each Model

--> For each model steps to follow;

• Model and Prediction
• Evaluation of Model
• Model Tuning
• Model Visualization (Feature Importances, ROC/AUC Curve, Confusion Matrix, etc.)
• Saving the Model