Machine Learning Diabetes Prediction

Algorithms Used:

1. Logistic Regression
2. Support Vector Classifier
3. KNearest Neighbors Classifier
4. Random Forest Classifier

Python Libraries Used:

1. Numpy
2. Pandas
3. Matplotlib
4. Seaborn
5. Scikit-learn

Exploratory Data Analysis and Visualization Summary:

• There are no missing values in the dataset
• The dataset is imbalanced
• No negative and closer to 1.0 correlations based on the correlation matrix
• The features are skewed and have outliers

Conclusion (Comments):

• It seems that the SVC model has reached its maximum potential on a imbalanced dataset
• Other solutions is to penalize or apply regularization and gradient descent techniques in the selected models
• Other solutions is to have more data to solve the imbalanced dataset or use sampling techniques
• After that, select a few more classification models based on the defined problem, type of data, and the expected outcome

Noticed Mistakes:

• Scaled all the features too early
• Fit only for the x_train and tranform only for the x_test dataset using the StandardScaler