Exploiting ChatGPT generated dataset for weather prediction using Seagull Optimization with Deep Learning Model

Abstract

ChatGPT, a powerful language model, can generate diverse fictional weather datasets including temperature, humidity, wind speed, and more. This research proposes the SGOADL-WDC technique, which employs ChatGPT-generated data for weather classification. It integrates data preprocessing with an attention-based bidirectional LSTM (ABiLSTM) for effective pattern recognition. The model’s performance is further optimized using the Seagull Optimization Algorithm (SGOA). By leveraging ChatGPT for synthetic data generation and deep learning for analysis, the SGOADL-WDC method demonstrates superior accuracy in classifying weather conditions, offering a novel approach to weather prediction compared to traditional methods reliant on historical data and numerical simulations.