https://www.sabapub.com/index.php/jaai/issue/feed 2023-04-20T22:53:06+00:00 Open Journal Systems <p>Journal of Applied Artificial Intelligence (JAAI) is an international and interdisciplinary scholarly peer reviewed journal on artificial intelligence published by Saba Publishing.<br />JAAI devoted entirely to Artificial Intelligence and welcomes papers in the overall field including, but not limited to, machine learning and cognition, deep learning, supervised learning, unsupervised learning, classification, regression, clustering, big and streaming data, optimization algorithms, feature selection and extraction, pattern recognition, bio-informatics, uncertain information processes, recommender systems, E-service personalization, distributed and parallel processing, computer vision, neural networks, natural language processing, heuristic search, multi-objective optimization, multi-agent systems, advances in social network systems, reasoning under uncertainty, forecasting and predication models as well as other hot topics.</p> <p><strong>Editor in Chief: <a href="https://www.scopus.com/authid/detail.uri?authorId=55497463100" target="_blank" rel="noopener">Dr Nibras Abdullah</a></strong><br /><strong>ISSN (online)</strong>: <a href="https://portal.issn.org/resource/ISSN/2709-5908" target="_blank" rel="noopener">2709-5908</a><br /><strong>Frequency:</strong> Semiannual</p> https://www.sabapub.com/index.php/jaai/article/view/692 2023-02-06T16:10:20+00:00 Hana Amar hanaomarkanan1@gmail.com

<p>Ensemble learning is a powerful technique for constructing accurate predictive models. Feature subset generation is an important step for ensemble learning. This paper proposes feature clustering and mutual information as a new feature subset generation method for ensemble learning. The proposed feature subset generation technique clusters the features using a hierarchical clustering algorithm. Mutual information is used to compute the similarity between the features within each cluster. Feature subset generation is then performed by selecting the most informative features from each cluster. Experiments are conducted on a real-world dataset to compare the proposed feature subset generation technique to other existing feature subset generation techniques. The experimental results show that the proposed technique outperforms existing feature subset generation techniques.</p> <p>&nbsp;</p>

2023-04-20T00:00:00+00:00 Copyright (c) 2023 Journal of Applied Artificial Intelligence