Knowledge discovery in weather forecasting: mining fuzzy image association rules with fine-tuned CNN and fuzzy HIFP algorithm

In this paper, a novel approach for mining image association rules is presented, which involves the fine-tuned VGG-16 model, proposed fuzzy image transactional database, as well as the proposed Fuzzy HIFP algorithm. In the first phase, fuzzy optimized image transactional database is generated using feature vectors obtained from the fine-tuned VGG-16 model. In the second phase, the proposed Fuzzy HIFP algorithm is used to generate fuzzy optimized image association rules. This proposed algorithm makes use of a hybrid approach that combines the effectiveness of fuzzy frequent pattern growth and hash indexing techniques. Experiments are performed on a weather image dataset consisting of 1250 images from four different classes, and the results shows that the proposed methodology outperforms existing approaches in terms of image association rules generation which can be further extended for classification purposes. Three runs are conducted in order to determine the optimized value of support and confidence. The obtained fuzzy image association rules provide valuable insights into the relationships and dependencies among the weather image features. These findings can support decision-making, pattern recognition, and knowledge discovery in weather forecasting and related domains.