An MCDM Approach for Enhancing Breast Cancer Diagnosis Using AHP and Bidirectional RNN with Bayesian Optimization

On average, one woman loses her life to breast cancer every minute. On the other hand, there are currently an unprecedented number of grounds for optimism. Breast cancer patients have a greater surviving chance if they obtain an early diagnosis of the disease. In order to conduct a more accurate assessment of machine learning models, the purpose of this research is to make use of an innovative strategy that integrates artificial intelligence (AI) with a multi-criteria decision-making (MCDM) process. The MCDM technique[1] that has been implemented incorporates the Preference AHP technique for Ranking Evaluations. The neural network model considered is the Bidirectional Recurrent Neural Network and for hyper-parameter tuning and optimization of the performance, Bayesian Optimization is considered. The proposed model consisting of BRNN and BES is found to be the most advantageous model for the early identification of breast cancer. It was able to provide a performance accuracy of 99.5%