The Tabu_Genetic Algorithm: A Novel Method for Hyper-Parameter Optimization of Learning Algorithms

Abstract

Machine learning algorithms have been widely used to deal with a variety of practical problems such as computer vision and speech processing. But the performance of machine learning algorithms is primarily affected by their hyper-parameters, as without good hyper-parameter values the performance of these algorithms will be very poor. Unfortunately, for complex machine learning models like deep neural networks, it is very difficult to determine their hyper-parameters. Therefore, it is of great significance to develop an efficient algorithm for hyper-parameter automatic optimization. In this paper, a novel hyper-parameter optimization methodology is presented to combine the advantages of a Genetic Algorithm and Tabu Search to achieve the efficient search for hyper-parameters of learning algorithms. This method is defined as the Tabu_Genetic Algorithm. In order to verify the performance of the proposed algorithm, two sets of contrast experiments are conducted. The Tabu_Genetic Algorithm and other four methods are simultaneously used to search for good values of hyper-parameters of deep convolutional neural networks. Experimental results show that, compared to Random Search and Bayesian optimization methods, the proposed Tabu_Genetic Algorithm finds a better model in less time. Whether in a low-dimensional or high-dimensional space, the Tabu_Genetic Algorithm has better search capabilities as an effective method for finding the hyper-parameters of learning algorithms. The presented method in this paper provides a new solution for solving the hyper-parameters optimization problem of complex machine learning models, which will provide machine learning algorithms with better performance when solving practical problems.

Optimal Planning Method for Large-Scale Historical Exhibits in the Taiwan Railway Museum

“…A genetic algorithm (GA) is used to find an optimal solution or suboptimal solution to a difficult optimization problem [43,44]. A GA is a machine learning (ML) algorithm that reflects the process of natural selection, where the fittest individuals are selected for reproduction in order to produce offspring of the next generation [33,[43][44][45] (Figure 3). The GA operates with a set of problem solutions, referred to as a population.…”

“…A subset of individuals is selected as the parents based on their high fitness value. The next generation is obtained thanks to crossover and mutation [33,43,44]. The gene, chromosome, and population, in this paper, means the train, layout instance, and layout instances separately.…”

“…It is a good way to refer to and modify the equation from Guo et el. [43]-see Equation 2, where A is a learning algorithm and λ is the type score of A. The ultimate goal of the optimization is to find the optimal total scores λ ( * ) to maximize the generalization score G χ [V(χ; A λ (χ (train) ))] of the learning algorithm A on dataset χ (train) (a P (n, r) = n!…”

Swarm‐based optimizer for convolutional neural network: An application for flood susceptibility mapping

“…In contrast, the implementation of optimization algorithms (as a replacement for gradient descents) in searching for optimal kernel values and weights of a CNN is a primary challenge (Guo, Hu, Wu, Peng, & Wu, 2019) and is still rare (Rere, Fanany, & Arymurthy, 2016). Some studies proposed and verified potential uses of meta‐heuristic approaches in science, engineering, and industries, such as the use of generic and Tabu search algorithms (Guo et al, 2019), micro‐canonical annealing (Ayumi, Rere, Fanany, & Arymurthy, 2016), or simulated annealing, evolution, and harmony search in the study of Rere et al (2016). The results showed improvements in classification accuracies and smaller errors, even though there was a computational time increase (Guo et al, 2019; Rere et al, 2016).…”

An Approach to Hyperparameter Optimization for the Objective Function in Machine Learning

“…In order to calculate the most appropriate learning rate with a minimum preliminary result value, a Tree-structured Parzen Estimator (TPE) [6] has been studied. Regarding optimization, techniques using Taub search [7] or other methods [8,9] have been presented. Bayesian optimization estimates the parameter distribution using prior values.…”