Hybrid PSO-HHO Based Machine Learning Optimization For Lung Cancer Diagnosis

An increased mortality rate is related to lung cancer. It is amongst the primary category of cancer. Lung cancer is a tumor that develops in the lung tissues, specifically in the linings of respiratory passages. Several factors contribute to causing lung cancer. Smoking, air pollution, and industrial exposure are a few of them. Survival rate deviations are subjective to the stage of disease progression at the time of detection, highlighting the importance of timely screening. This study utilized a publicly available data set and applied a hybrid PSO-HHO algorithm for feature selection. It is an important step that discards irrelevant features and enhances the capability of machine learning models. Six machine learning algorithms, Gaussian Naive Bayes, Decision Tree, K Nearest Neighbors, Multinomial Naive Bayes,Logistic Regression, and Random Forest are employed for performance evaluation of the obtained feature subset from the hybridized model. It is observed that Hybrid PSO-HHO, with Logistic Regression outperformed other algorithms with 92.59 % accuracy. In the healthcare diagnostic system, the suggested model can significantly contribute to efficient treatment and an increased survival rate.