One of the main causes of death in the world today is cancer. Cancer is caused by a number of reasons, but early identi-fication and precise prediction at the appropriate stage are essential. The main goal of this study is to compare several machine learning techniques for early cancer prediction to the Genetic Algorithm (GA) for early-stage cancer detection. The Genetic Algorithm modifies the link weights to optimize neural networks. Several methods, such as Random Forest, Naïve Bayes, Artificial Neural Network (ANN), and Genetic Algorithm, are trained and tested using the Breast Cancer Wisconsin (Original) dataset, which is obtained from the UCI Machine Learning Repository. To attain high accuracy and dependable performance, the cancer microarray data is subjected to Random Forest, an ensemble learning technique. The UCI breast cancer dataset is also utilized for predictive modelling using the weight-assignment-based Naïve Bayes method. By using the backpropagation algorithm, the Artificial Neural Network reduces the discrepancy between predicted and actual results by evaluating output by calculating and modifying mistakes through the use of numerous neurons in the hidden layer. Experimental analysis shows that the Genetic Algorithm yields the highest accuracy at 93%, while other algorithms such as Artificial Neural Network, Naïve Bayes, and Random Forest provide accuracies of 92%, 94%, and 96%, respectively.

Cancer, Genetic, Neural Network, Random Forest