AI (Artificial intelligence) developments in recent years have converted the traditional healthcare system into smart healthcare. Medical services may be enhanced by embracing important technology like AI. The healthcare sector has a variety of opportunities because to AI convergence. Moreover, data mining has been crucial in revealing hidden patterns in huge datasets. The current method has been shown to have issues with reduced classification accuracy in earlier studies. In this study, AFOECNN (Adaptive Firefly Optimization with Enhanced Convolution Neural Networks) is developed to increase the classification accuracy in order to address the aforementioned issue. Pre-processing, choosing feature subsets, and classification are the three primary stages of the proposed approach. KMC (K-Means Clustering) technique is used for pre-processing in order to reduce the noise data from the provided gene expression dataset. By employing k-means centroid values to manage missing features, it more effectively raises classification accuracy. The pre-processed features are employed in the feature subset selection approach to extract more beneficial features from the cancer dataset. Based on the best fitness values, the vital and obvious characteristic is computed using the objective function. It is carried out using the AFO (Adaptive Firefly Optimisation) algorithm. The data is transformed into pictures, and then augmentation is applied to those images. Finally, using a training and testing model, the ECNN algorithm is employed for classification. Using weight values, it more correctly categorises the tumour traits. This work’s experimental ioutcomes show that AFOECNN is better than existing algorithms in terms of precision, recall, f-measure, accuracy values and reduced time complexities.

Gene expression data, Artificial Intelligence (AI), data mining, Adaptive Firefly Optimization with Enhanced Convolution Neural Network (AFOECNN) algorithm