Hyperparameter tuning of the model for hunger state classification

• Main Authors: Razman, Mohd Azraan, Abdul Majeed, Anwar P.P., Musa, Rabiu Muazu, Taha, Zahari, Susto, Gian Antonio, Mukai, Yukinori
• Format: Book Chapter
• Language: English; English
• Published: Springer 2020
• Subjects: SH Aquaculture. Fisheries. Angling; SH151 Aquaculture - Fish Culture
• Online Access: http://irep.iium.edu.my/83297/1/83297_Hyperparameter%20tuning%20of%20the%20model_MYRA.pdf; http://irep.iium.edu.my/83297/2/83297_Hyperparameter%20tuning%20of%20the%20model_SCOPUS.pdf; http://irep.iium.edu.my/83297/; https://doi.org/10.1007/978-981-15-2237-6_5

Abstract To increase the classification, the rate of prediction based on existing 2 models requires additional technique or in this case optimizing the model. Hyper3 parameter tuning is an optimization technique that evaluates and adjusts the free 4 parameters that define the behaviour of classifiers. Data sets were classified practical 5 with classifiers like SVM, k-NN, ANN and DA. To further improve the design effi6 ciency, the secondary optimization level called hyperparameter tuning will be further 7 investigated. DA, SVM, k-NN, decision tree (Tree), logistic regression (LR), random 8 forest tree (RF) and neural network (NN) are evaluated. The k-NN provided 96.47% 9 of the test sets with the best reliability in classifications. Bayesian optimization has 10 been used to refine the hyperparameter; hence, standardize Euclidean distancemetric 11 with a k value of one is the ideal hyperparameters which could achieve classification 12 performance of 97.16%.