Lung Cancer Prediction and Recommendations Using Machine Learning and Generative AI

Lung Cancer Prediction and Recommendations Using Machine Learning and Generative AI

Abhishek Ponnaboina1, Dr. S. China Venkateswarlu2, Dr. V. Siva Nagaraju3, Dr. Prashant Bachanna4

1Department of Electronics and Communication Engineering Institute of Aeronautical Engineering, Hyderabad, INDIA abhi04457@gmail.com

2Professor, Department of Electronics and Communication Engineering

Institute of Aeronautical Engineering, Hyderabad, INDIA c.venkateshwarlu@iare.ac.in

3Professor, Department of Electronics and Communication Engineering

Institute of Aeronautical Engineering, Hyderabad, INDIA v.sivanagaraju@iare.ac.in

4Professor, Department of Electronics and Communication Engineering

Institute of Aeronautical Engineering, Hyderabad, INDIA b.prashant@iare.ac.in

ABSTRACT
This project presents an AI-driven web application designed to predict lung cancer risk and provide person- alized health recommendations. Utilizing a dataset with multiple patient health indicators, the system employs data preprocessing techniques such as categorical encoding and feature scaling. To address class imbalance, Synthetic Minority Oversampling Technique (SMOTE) is applied. The predictive model is built using a Random Forest Classifier trained on the balanced and scaled data, achiev- ing reliable performance for lung cancer detection. The application is implemented with Streamlit, enabling users to input health parameters and receive instant diagnostic results. Additionally, Google Gemini’s generative AI model is integrated to generate concise, doctor-like explanations, outlining causes, symptoms, lifestyle risks, and tailored recommendations based on the prediction. This approach demonstrates the potential of combining machine learning and generative AI to enhance early lung cancer detection and improve preventive healthcare outcomes.
Keywords—Lung cancer prediction, machine learning, Ran- dom Forest, SMOTE, generative AI, healthcare analytics, Stream- lit, supervised learning.