An Intelligent Scheduling Framework for AI-driven Geographic Simulation in Distributed Heterogeneous Environments Experiment Data and Results

These files are the model input data and experimental results of the manuscript entitled "An Intelligent Scheduling Framework for AI-driven Geographic Simulation in Distributed Heterogeneous Environment" submitted to GIScience&remote sensing. The "UrbanM2M_model_input_data.zip" is the input data of urbanM2M model, which includes the input data of three research areas around Taihu Lake, Suzhou and Kunshan; The "UrbanVCA_model_input_data.zip" is the input data of urbanVCA model, which includes the input data of two research areas around Shunde and Shanghua; The "Vision-LSTM_model_input_data.zip" is the input data of Vision-LSTM model, which includes 289, 578 and 1186 test data of different quantities. The "Comparative analysis of task allocation and execution performance for eight geo-simulation tasks between current and AI scheduling strategies.xlsx" is the result data of Experiment 1 "Decision Process Analysis for Individual Task"; The "Comparative performance analysis of task throughput, success rates, and total runtimes under current static and AI dynamic scheduling strategies for three different geo-simulation models.xlsx" is the result data of Experiment 2 "Performance Comparison under Equal Task Volumes"; The "Comparison of task execution outcomes and scheduling stability under current static scheduling strategy and AI dynamic scheduling strategy over 4 hours.xlsx" is the result data of Experiment 3 "Throughput Evaluation within Equal Time Frames". The data shall not be transmitted or used without the consent of the author.