Efficient Model for Prediction of Non-small Cells Lung Cancer via Deep Q-Learning

Accurate prognosis is crucial for fast detection of non-small cell lung cancer (NSCLC). We address the need for a powerful deep Q-learning feature representation model for NSCLC prediction in this paper. We propose a novel approach for the segmentation of non-small cells, combining adaptive thresholding with a number of feature extraction techniques, including Fourier, Cosine, Wavelet, Laplace, Z Transform, S Transform, and Gabor Transform. These techniques are utilized to extract distinctive traits from the input data, enabling a more precise and comprehensive description of NSCLC features. To address these problems, we present a 1D convolutional neural network (CNN) incorporating Q-learning, a reinforcement learning technique that allows for ongoing accuracy improvements. The effectiveness of the core algorithms in our proposed model makes it stand out. Proposed method finds 12.5% more precision.