Predicting Diabetes Using Machine Learning – A Medium-Level ML Project

This project demonstrates how to predict whether a patient has diabetes using machine learning. We use the Pima Indians Diabetes dataset and walk through a full ML workflow including data cleaning, model building, and evaluation.

Tools Required: Python, pandas, scikit-learn, matplotlib, seaborn

Step 1: Import Required Libraries

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import classification_report, confusion_matrix, accuracy_score

Step 2: Load the Dataset

Download the dataset from Kaggle: Pima Indians Diabetes Dataset

df = pd.read_csv("diabetes.csv")
df.head()

Step 3: Explore and Clean the Data

df.info()
df.describe()

Replace invalid 0 values in key health-related columns:

cols = ['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI']
df[cols] = df[cols].replace(0, np.nan)
df.fillna(df.mean(), inplace=True)

Step 4: Visualize the Data

sns.countplot(x='Outcome', data=df)
plt.title("Diabetes Count (0 = No, 1 = Yes)")
plt.show()

sns.pairplot(df, hue='Outcome')
plt.show()

Step 5: Split and Scale the Data

X = df.drop('Outcome', axis=1)
y = df['Outcome']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

scaler = StandardScaler()
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)

Step 6: Train the Model

model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

Step 7: Make Predictions and Evaluate

y_pred = model.predict(X_test)

print("Accuracy:", accuracy_score(y_test, y_pred))
print("Classification Report:\n", classification_report(y_test, y_pred))

# Confusion matrix
cm = confusion_matrix(y_test, y_pred)
sns.heatmap(cm, annot=True, fmt='d', cmap='Blues')
plt.xlabel('Predicted')
plt.ylabel('Actual')
plt.title('Confusion Matrix')
plt.show()

Step 8: Feature Importance

importances = model.feature_importances_
features = df.columns[:-1]

plt.barh(features, importances)
plt.xlabel("Feature Importance")
plt.title("Important Features in Prediction")
plt.show()

Step 9: Conclusion

• Random Forest gave strong accuracy on the diabetes prediction task.
• Glucose, BMI, and Age were highly important features.
• This workflow can be adapted to other health or classification problems.

What's Next?

• Try logistic regression or XGBoost for comparison.
• Tune hyperparameters using GridSearchCV.
• Deploy this model using Streamlit or Flask for doctors to use!