Grey-Topsis and Optimized RNN Algorithms with Dynamic Weight Adjustment: Construction of a Precise Prediction Model for the Global Cybercrime Situation
Yuyang Dai, Runze Yuan, Zhen Tian · 2025
This paper proposes a hybrid model integrating Grey-Topsis and Recurrent Neural Network (RNN) for predicting the global cybercrime trend. Data from 38 countries from 2010 to 2024 were collected, and after data preprocessing and CRITIC-based weighting, an improved Grey-TOPSIS evaluation system was applied. The RNN model was optimized with dynamic weight adjustment. K-means clustering was used to analyze crime patterns, and the DID method was employed to verify the effectiveness of policy interventions. The experimental results show that the average relative error of the predictions for 2025 - 2030 is 17.72%, providing strong support for formulating targeted strategies.