HYPER-PARAMETER OPTIMIZATION AND EVALUATION ON SELECTED MACHINE LEARNING ALGORITHM USING HEPATITIS DATASET

  • Aminat Yusuf UDUS
  • Oyelola Akande
Keywords: Hepatitis disease, Grid Search, Parameters Tuning, Machine Learning Techniques

Abstract

Despite the popularity and utility of most machine learning techniques, expert knowledge is required in guiding choices about the suitable technique and settings that are good for solving a specific problem. The lack of expert information renders the procedures vulnerable to poor parameter settings. Several of these machine learning techniques configurations are offered under default settings. However, since different classification problems required suitable machine learning techniques, selecting the appropriate technique and tuning its settings are vital works that will rightly improve predictions in terms of reliability and accuracy. This study aims to perform grid search parameters tuning on 5-selected machine learning techniques on hepatitis disease. Comparative performance is drawn side-by-side with the default settings. The experimental results of the five tuning techniques show that using the configurations suggested in our work yield predictions of a greatly sophisticated quality than choice under its default settings. The result proves that tuning parameters of Support Vector Machine via grid search yields the best accuracy outcomes of 90% and has a competitive performance relative towards criteria of precision, recall, accuracy and Area Under the Curve. Present combinations of parameter settings for each of the techniques by identifying ranges of values for each setting that give good Hepatitis disease outcomes

References

Bermúdez-Chacón, R., Gonnet, G. H., & Smith, K. (2015). Automatic problem-specific hyperparameter optimization and model selection for supervised machine learning: Technical Report: Technical Report (p. 52 p.) [Application/pdf]. ETH Zurich. https://doi.org/10.3929/ETHZ-A-010558061

Bhargav, K. S., & Kumari, T. D. (2018). Application of Machine Learning Classification Algorithms on Hepatitis Dataset. 13(16), 6.

Cequera, A., & García, L. M. (2014). Biomarkers for liver fibrosis: Advances, advantages and disadvantages. Revista de Gastroenterología de México (English Edition), 79(3), 187–199. https://doi.org/10.1016/j.rgmxen.2014.07.001

Chown, H. (2019). A Comparison of Machine Learning Algorithms for the Prediction of Hepatitis C NS3 Protease Cleavage Sites. Journal of Proteomics & Bioinformatics, 12(5). https://doi.org/10.35248/0974-276X.19.12.501

Parisi. (2014). Exploiting Kinetic and Kinematic Data to Plot Cyclograms for Managing the Rehabilitation Process of BKAs by Applying Neural Networks. 8(10), 5.

Parisi, Biggs, P. R., Whatling, G., & Holt, C. A. (2015). A Novel Comparison of Artificial Intelligence Methods for Diagnosing Knee Osteoarthritis. https://doi.org/10.13140/RG.2.1.4197.6163

Ramadhan, M. M., Sitanggang, I. S., Nasution, F. R., & Ghifari, A. (2017). Parameter Tuning in Random Forest Based on Grid Search Method for Gender Classification Based on Voice Frequency. 6.

Yarasuri, V. K., Indukuri, G. K., & Nair, A. K. (2019). Prediction of Hepatitis Disease Using Machine Learnin
Published
2021-07-13
How to Cite
Yusuf, A., & Akande, O. (2021). HYPER-PARAMETER OPTIMIZATION AND EVALUATION ON SELECTED MACHINE LEARNING ALGORITHM USING HEPATITIS DATASET . FUDMA JOURNAL OF SCIENCES, 5(2), 447 - 455. https://doi.org/10.33003/fjs-2021-0502-649