Early Prediction of Gastric Cancer Disorders Utilising Random Forest and Patient Lifestyle Information

The current research focuses on the early identification of gastric cancer health-risk by using the Random Forest supervisor learning algorithm merged with patient lifestyle information’s. Gastric cancer is a leading cause of mortality worldwide, and early identification is important for improving survival rates. Traditional diagnostic methodology often fails to identify the disease at an early stage. For addressing this gap, the proposed framework integrates patient-specific lifestyle data like dietary habits, smoking frequency, alcohol intake, and daily physical activity—with a supervisor learning model to predict the likelihood of developing gastric cancer. The Random Forest Schema is used due to its robust capability to handle higher-dimensional data and capture complex relationships between lifestyle factors and cancer health risk. This model aims to assist healthcare experts in identifying higher-risk individuals, enabling timely interventions and any personalized treatment strategies. The proposed system uses the potential to improvise the effectiveness of gastric cancer screening and risk assessment depending on a combination of predictive analytics and patient behavior. The analysis results shows that Random Forest model outperforms other models in terms of accuracy (92%) and AUC-ROC (93%), while Neural Networks follow closely with an accuracy of 91%. Support Vector Machine and Logistic Regression have the lowest performance across all metrics.