In the era of Industrial 4.0, many urgent issues in the industries can be effectively solved with artificial intelligence techniques, including machine learning. Designing an effective machine learning model for prediction and classification problems is an ongoing endeavor. Besides that, time and expertise are important factors that are needed to tailor the model to a specific issue, such as the green building housing issue. Green building is known as a potential approach to increase the efficiency of the building. To the best of our knowledge, there is still no implementation of machine learning model on GB valuation factors for building price prediction compared to conventional building development. This paper provides a report of an empirical study that model building price prediction based on green building and other common determinants. The experiments used five common machine learning algorithms namely Linear Regression, Decision Tree, Random Forest, Ridge and Lasso tested on a set of real building datasets that covered Kuala Lumpur District, Malaysia. The result showed that the Random Forest algorithm outperforms the other four algorithms on the tested dataset and the green building determinant has contributed some promising effects to the model.

Student acceptance of educational games in higher education

This paper addresses the performances of machine learning classification models for the detection of tax avoidance problems. The machine learning models employed automated features selection with hybrid two metaheuristics algorithms namely particle swarm optimization (PSO) and genetic algorithm (GA). Dealing with a real dataset on the tax avoidance cases among companies in Malaysia, has created a stumbling block for the conventional machine learning models to achieve higher accuracy in the detection process as the associations among all of the features in the datasets are extremely low. This paper presents a hybrid meta-heuristic between PSO and adaptive GA operators for the optimization of features selection in the machine learning models. The hybrid PSO-GA has been designed to employ three adaptive GA operators hence three groups of features selection will be generated. The three groups of features selection were used in random forest (RF), k-nearest neighbor (k-NN), and support vector machine (SVM). The results showed that most models that used PSO-GA hybrids have achieved better accuracy than the conventional approach (using all features from the dataset). The most accurate machine learning model was SVM, which used a PSO-GA hybrid with adaptive GA mutation.

Particle Swarm Optimization (PSO) is a popular algorithm used extensively in continuous optimization. One of its well-known drawbacks is its propensity for premature convergence. Many techniques have been proposed for alleviating this problem. One of the alternative approaches is hybridization. Genetic Algorithms (GA) are one of the possible techniques used for hybridization. Most often, a mutation scheme is added to the PSO, but some applications of crossover have been added more recently. Some of these schemes use adaptive parameterization when applying the GA operators. In this work, adaptively parameterized mutation and crossover operators are combined with a PSO implementation individually and in combination to test the effectiveness of these additions. The results indicate that an adaptive approach with position factor is more effective for the proposed PSO hybrids. Compared to single PSO with adaptive inertia weight, all the PSO hybrids with adaptive probability have shown satisfactory performance in generating near-optimal solutions for all tested functions.

Machine learning has been very promising in solving real problems, but the implementation involved difficulties mainly for the inexpert data scientists. Therefore, this paper presents an automated machine learning (AutoML) to simplify and accelerate the modeling tasks. Focused on Python and RapidMiner rapid modeling tools, Tree-based Pipeline Optimization Tool (TPOT) and AutoModel were used. This paper presents a comprehensive comparison between these tools with regard to the prediction accuracy and Area Under Curve (AUC) in classifying real cases of whistleblowing academic dishonesty among undergraduate students of two universities in Indonesia. Additionally, the correlations weight from demographic and Theory of Planned Behavior (TOB) attributes in the different machine learning models are also discussed. All the machine learning algorithms from TPOT and AutoModel are considerable powerful to generate good accuracy level (between 70–93% of AUC) in classifying both cases of whistleblowing and non-whistleblowing on the hold-out samples from the testing process. Generally, based on the validation results of the prediction models, demographic attributes presented more importance than the TBP attributes. The findings of this study will be a great interest of many research scholars to conduct a more in-depth analysis on AutoML for many domains mainly in education and academic misconduct fields. • AutoML is the first of its kind to be empirically compared between TPOT and AutoModel in an application to predict academic dishonesty whistleblowing. • Besides accuracy performances of the AutoML, the proportion of the variance of each attribute from demographic and Theory of Planned Behavior (TPB) is also presented in the prediction models of academic dishonesty whistleblowing. • AutoML is a convenient and reproducible rapid modeling method of machine learning to be used in many kinds of prediction problem.

Designing an effective machine learning model for prediction or classification problem is a tedious endeavor. Significant time and expertise are needed to customize the model for a specific problem. A significant way to reduce the complicated design is by using Automated Machine Learning (AML) that can intelligently optimize the best pipeline suitable for a problem or dataset. This paper demonstrates the utilization of an AML that has been developed with a meta-heuristic algorithm namely Genetic Programming (GP). Empirical experiment has been conducted to test the performances of AML on a real dataset of house prices in the area of Petaling Jaya, Selangor. The results show that the AML with GP able to produce the best pipeline of machine learning with high score of accuracy and minimal error. (Abstract)