| Abstrakt: |
In recent years, metaheuristic optimizers have grown in popularity due to their ability to efficiently optimize complex, high-dimensional problems that are difficult to solve using conventional optimization techniques. In the context of multiple learning models, where large-scale datasets and complex models necessitate efficient optimization techniques, these optimization algorithms are particularly useful. The objective of this paper is to design a system that combines metaheuristic optimizers with various AI based classifiers to detect and diagnose skin diseases. In order to accomplish this objective, numerical and image datasets have been taken, pre-processed, and visually analysed in order to comprehend their patterns. Features have been extracted by computing different characteristics of an image such as mean intensity, area, aspect ratio, width, etc. To classify different skin diseases, multiple learning models are incorporated with the numbers of hyperparameter optimizers and it has been observed that the decision tree model achieved 87.30% accuracy under random search CV. Under Bayesian search cv, the deep neural network model computed the highest accuracy of 90.56%, whereas ResNet50 achieved 74.89% accuracy without tuning the hyperparameter. These results indicate that incorporating hyperparameter metaheuristic techniques is able to improve the detection as well as diagnosis performance of models for skin disease. However, additional research is required to investigate new strategies and enhance the scalability of the current methods. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink