Detection of Potential Biomarkers in Lung Adenocarcinoma Using WCC Algorithm and Futuristic Greedy index with Feature Selection Approach

Lung cancer is the second-leading cause of death worldwide. One method of cancer prediction involves identifying genes that are effective in cancer. The selection of efficient genes is one area of interest for specialists in diagnosing lung cancer. This research presents a method based on mathematical modelling that considers genes with high dimensions. The output of solving the mathematical model is to determine efficient genes for lung cancer diagnosis. We first remove ineffective genes in lung cancer diagnosis using the proposed method. Then, for the remaining genes, we obtain the dependency matrix between them. In the next step, we get the sorted solution space for the World Competitive Contests algorithm (WCC) using the futuristic greedy index and the dependency matrix. The research's main innovation is identifying this ordered solution space within the larger solution space. The ordered solution space is a space of candidate solutions not randomly selected by the WCC algorithm. Instead, these candidate solutions are regularly used by the WCC algorithm to find proper approximate solutions. This method reduces the number of iterations in the WCC algorithm and finds appropriate approximate solutions in less execution time. The experimental results showed that in the first dataset, the proposed method achieved a 17.63% improvement in execution time compared to the fast WCC algorithm. Additionally, with the increase in data volume in the large dataset, the execution time of the proposed algorithm shows an improvement of approximately 41.6% compared to the WCC algorithm. Experiments confirm that the proposed method's Accuracy error rate is approximately 1%, and that the error rate varies significantly as the data sets change and grow, because the WCC algorithm prevents random choices in the solution space.