Hybrid ANN-Based Model for Lung Cancer Disease Prediction

Currently, lung cancer ranks as the world's leading cause of cancer-related fatalities in both men and women. The primary cause of lung cancer may be attributed to smoking. Although lung cancer can develop in any part of the lung, 90% to 95% of cases are believed to start in the epithelial cells that line the larger and smaller airways (bronchi and bronchioles). This study's main goal is to identify lung cancer by employing artificial neural networks with scaled conjugate gradient back propagation and k-fold cross validation. We have utilized the Lung Cancer dataset through kaggle. Data from 1000 lung cancer patients, ages 14 to 73, is used to train the network. How many neurons are in the hidden layer affects how accurate the results are. The proposed system uses 11 input attributes to assess the presence and absence of lung cancer during testing, yielding a minimum accuracy percentage of 83.1429% and a maximum accuracy percentage of 100%.