Hybrid Deep Neural Networks and Genetic Algorithms for Stock Forecasting with Decentralized Blockchain

This research introduces a hybrid stock market forecasting system which integrates LSTM–CNN deep learning, Meyer Packard Genetic Algorithm optimization as well as blockchain-based validation for transparent financial forecasts. Both long-term and short-term dependencies in stock market data are captured by our hybrid LSTM-CNN model, and the genetic algorithm optimizes model hyperparameters for better and improved forecasting accuracy. On the Ethereum Sepolia testnet, predictions and trading signals are stored using blockchain technology via Solidity smart contracts, producing an immutable and auditable record. The proposed system improves performance, transparency reliability when compared to conventional machine learning methods. Findings show improved accuracy and accountability, providing a solid foundation for next-generation financial forecasting [1, 3, 7].