Retinal degenerative diseases, including age-related macular degeneration (AMD) and retinitis pigmentosa (RP), are leading causes of irreversible vision loss. This study presents a comprehensive bioinformatics analysis of retinal gene expression using machine learning and deep learning methods. RNA-seq data from 284 retinal samples (152 patients with degenerative diseases and 132 controls) were analyzed. A convolutional neural network (CNN)-based model was developed for classifying pathological conditions with an accuracy of 94.3% (AUC = 0.97). A total of 847 differentially expressed genes (DEGs) with |log₂FC| > 1.5 and p < 0.001 were identified. Key signaling pathways included photoreceptor apoptosis (p = 2.3×10⁻¹²), oxidative stress (p = 4.7×10⁻¹⁰), and inflammation (p = 1.2×10⁻⁹). Application of the WGCNA algorithm identified 12 gene coexpression modules, of which 3 modules showed a strong correlation with clinical parameters (r > 0.75, p < 0.0001).