Optimizing Crop Recommendations with Improved Deep Belief Networks: A Multi model Approach

This study proposes a multi-model framework based on an Improved Deep Belief Network (IDBN). The system incorporates Gaussian Restricted Boltzmann Machines to effectively process continuous soil-related information and utilizes the Ranger optimizer to achieve faster and more stable convergence. The approach supports better resource planning and enhances sustainability by reducing the limitations of conventional, experience-driven decision methods. Multiple datasets obtained from diverse sources were cleaned, processed, and subjected to feature selection techniques to identify the most significant attributes. Using these refined features, the IDBN framework was developed and evaluated. Its performance was compared against several established machine-learning models to assess its effectiveness