Knowledge-Based Nonlinear to Linear Dataset Transformation for Chronic Illness Classification
Paper
Paper/Presentation Title | Knowledge-Based Nonlinear to Linear Dataset Transformation for Chronic Illness Classification |
---|---|
Presentation Type | Paper |
Authors | Jaworsky, Markian, Tao, Xiaohui, Yong, Jianming, Pan, Lei, Zhang, Ji and Pokhrel, Shiva Raj |
Journal or Proceedings Title | Proceedings of the 12th International Conference on Health Information Science (HIS 2023) |
Journal Citation | 14305, pp. 115-126 |
Number of Pages | 12 |
Year | 2023 |
Place of Publication | Singapore |
ISBN | 9789819971084 |
9789819971077 | |
Digital Object Identifier (DOI) | https://doi.org/10.1007/978-981-99-7108-4_10 |
Web Address (URL) of Paper | https://link.springer.com/chapter/10.1007/978-981-99-7108-4_10 |
Web Address (URL) of Conference Proceedings | https://link.springer.com/book/10.1007/978-981-99-7108-4 |
Conference/Event | 12th International Conference on Health Information Science (HIS 2023) |
Event Details | 12th International Conference on Health Information Science (HIS 2023) Parent International Conference on Health Information Science (HIS) Delivery In person Event Date 23 to end of 24 Oct 2023 Event Location Melbourne, Australia |
Abstract | Nonlinear patterns are challenging to interpret, validate, and are resource-intensive for deep learning (DL) and machine learning (ML) algorithms to predict chronic illness. Transformation of nonlinear features to a linear representation enables the human understanding of AI results and traditional and proven ML algorithms. We propose the counts of terms cross-checked against the chapters of the International Classification of Disease (ICD) to replace the raw representation of key nonlinear variables in health surveys to improve the chronic illness classification performance. The specific selection of nonlinear keywords viz. Male, Female, Diabetes, Cancer, Obese, Overweight, Smoked, Cigarettes, and Sugar from a health survey, transformed into a purely linear and scaled set of features propels the Multinomial Naive Bayes (MNB) algorithm to outperform standard dataset preparation and feature selection methods. |
Keywords | Risk Factors ; Linear Models; Nonlinear Models ; Chronic Illness; Knowledge Graphs |
Contains Sensitive Content | Does not contain sensitive content |
ANZSRC Field of Research 2020 | 460502. Data mining and knowledge discovery |
Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
Series | Lecture Notes in Computer Science |
Byline Affiliations | School of Mathematics, Physics and Computing |
School of Business | |
Deakin University |
https://research.usq.edu.au/item/z9w82/knowledge-based-nonlinear-to-linear-dataset-transformation-for-chronic-illness-classification
27
total views0
total downloads9
views this month0
downloads this month