| Abstrakt: |
The advances in the state-of-the-art methods addressing real-world AI applications vary from novel feature selection procedures [[7]] to the development of novel and application-based machine learning architectures [[1], [8]]. Recent advances in the field of artificial intelligence (AI) have been pivotal in enhancing the effectiveness and efficiency of many systems and in all fields of knowledge, including medical diagnosis [[1]], healthcare [[3]], the Internet of Things [[4]], power systems [[5]], etc. With the awareness of trustworthy AI and fairness, one contribution suggests the feature set extraction based on consistency of their explanation via different explainable AI methods (Contribution 9). Traditional machine learning models based on k-means and sequential minimal optimization methods (Contribution 23) and deep forest methods (Contribution 24) have been proposed for detecting anomalies and malicious traffic detection. [Extracted from the article] |
