Search results

Software defect prediction (SDP) is a critical aspect of software quality assurance, aiming to identify and manage potential defects in software systems. In this paper, we have proposed a novel hybrid approach that combines Gray Wolf Optimization with Feature Selection (GWOFS) and multilayer perceptron (MLP) for SDP. The GWOFS-MLP hybrid model is designed to optimize feature selection, ultimately enhancing the accuracy and efficiency of SDP. Gray Wolf Optimization, inspired by the social hierarchy and hunting behavior of gray wolves, is employed to select a subset of relevant features from an extensive pool of potential predictors. This study investigates the key challenges that traditional SDP approaches encounter and proposes promising solutions to overcome time complexity and the curse of the dimensionality reduction problem.

The integration of GWOFS and MLP results in a robust hybrid model that can adapt to diverse software datasets. This feature selection process harnesses the cooperative hunting behavior of wolves, allowing for the exploration of critical feature combinations. The selected features are then fed into an MLP, a powerful artificial neural network (ANN) known for its capability to learn intricate patterns within software metrics. MLP serves as the predictive engine, utilizing the curated feature set to model and classify software defects accurately.

The performance evaluation of the GWOFS-MLP hybrid model on a real-world software defect dataset demonstrates its effectiveness. The model achieves a remarkable training accuracy of 97.69% and a testing accuracy of 97.99%. Additionally, the receiver operating characteristic area under the curve (ROC-AUC) score of 0.89 highlights the model’s ability to discriminate between defective and defect-free software components.

Experimental implementations using machine learning-based techniques with feature reduction are conducted to validate the proposed solutions. The goal is to enhance SDP’s accuracy, relevance and efficiency, ultimately improving software quality assurance processes. The confusion matrix further illustrates the model’s performance, with only a small number of false positives and false negatives.

This study aims to investigate whether gender predicts financial inclusion and whether education can fill the gender gap in financial inclusion when controlling for the effects of supply side factors of financial inclusion in low-income economies.

This study aims to investigate whether gender predicts financial inclusion and whether education can fill the gender gap in financial inclusion when controlling for the effects of supply side factors of financial inclusion in low-income economies.

The findings provided support for the gender gap in financial inclusion using the most basic measure of financial inclusion. However, using formal savings and access to credit, the gender gap hypothesis is not supported. Moreover, the results revealed that education reduces the gender gap in the basic form of financial inclusion. However, this study could not find any significant difference between men and women's financial inclusion in terms of saving at a bank or borrowing from a bank though men tend to save more than women informally.

The current study contributes to the literature by examining the role of education in the relationship between gender gap and financial inclusion when controlling for the effects of heterogeneous infrastructure and the supply side factors of financial inclusion among the selected countries.

This study leverages social network analysis (SNA) to visualise the way students interacted with online resources and uses the data obtained from SNA as features for supervised machine learning algorithms to predict whether a student will successfully complete a course.

The exploration and visualisation of the data were first carried out to gain a better understanding of the students, the course(s) each student was enrolled in and each course’s virtual learning resources. Following this, the construction of the social network graphs was performed to depict how each student behaved online before the degree centralities were computed for each of the nodes in a social network graph. Data pre-processing to assign labels based on the final result a student obtained in a course was then performed before we trained and tested models to predict which students did or did not graduate.

The study’s findings demonstrate that the constructed predictive model has good performance, as shown by the accuracy, precision, recall and f-measure metrics. The outcomes also showed that students’ use of online resources is a crucial element that influences how well they perform in their academics.

The similarity index is as low as 9%.

With the demographic increase, especially in big cities, heavy traffic, traffic congestion, road accidents and augmented pollution levels hamper transportation networks. Finding the optimal routes in urban scenarios is very challenging since it should consider reducing traffic jams, optimizing travel time, decreasing fuel consumption and reducing pollution levels accordingly. In this regard, the authors propose an enhanced approach based on the Ant Colony algorithm that allows vehicle drivers to search for optimal routes in urban areas from different perspectives, such as shortness and rapidness.

An improved ant colony algorithm (ACO) is used to calculate the optimal routes in an urban road network by adopting an elitism strategy, a random search approach and a flexible pheromone deposit-evaporate mechanism. In addition, the authors make a trade-off between route length, travel time and congestion level.

Experimental tests show that the routes found using the proposed algorithm improved the quality of the results by 30% in comparison with the ACO algorithm. In addition, the authors maintain a level of accuracy between 0.9 and 0.95. Therefore, the overall cost of the found solutions decreased from 67 to 40. In addition, the experimental results demonstrate that the authors’ improved algorithm outperforms not only the original ACO algorithm but also popular meta-heuristic algorithms such as the genetic algorithm (GA) and particle swarm optimization (PSO) in terms of reducing travel costs and improving overall fitness value.

The proposed improvements to the ACO to search for optimal paths for urban roads include incorporating multiple factors, such as travel length, time and congestion level, into the route selection process. Furthermore, random search, elitism strategy and flexible pheromone updating rules are proposed to consider the dynamic changes in road network conditions and make the proposed approach more relevant and effective. These enhancements contribute to the originality of the authors’ work, and they have the potential to advance the field of traffic routing.

This study aims to investigate factors that could predict the continued usage of e-learning systems, such as the learning management systems (LMS) at a Technical University in Ghana using machine learning algorithms.

The proposed model for this study adopted a unified theory of acceptance and use of technology as a base model and incorporated the following constructs: availability of resources (AR), computer self-efficacy (CSE), perceived enjoyment (PE) and continuance intention to use (CIU). The study used an online questionnaire to collect data from 280 students of a Technical University in Ghana. The partial least square-structural equation model (PLS-SEM) method was used to determine the measurement model’s reliability and validity. Machine learning algorithms were used to determine the relationships among the constructs in the proposed research model.

The findings from the study confirmed that AR, CSE, PE, performance expectancy, effort expectancy and social influence predicted students’ continuance intention to use the LMS. In addition, CIU and facilitating conditions predicted the continuance use of the LMS.

The use of machine learning algorithms in e-learning systems literature has been rarely used. Thus, this study contributes to the literature on the continuance use of e-learning systems using machine learning algorithms. Furthermore, this study contributes to the literature on the continuance use of e-learning systems in developing countries, especially in a Ghanaian higher education context.

With the rapid growth of the domestic lending industry, assessing whether the borrower of each loan is at risk of default is a pressing issue for financial institutions. Although there are some models that can handle such problems well, there are still some shortcomings in some aspects. The purpose of this paper is to improve the accuracy of credit assessment models.

In this paper, three different stages are used to improve the classification performance of LSTM, so that financial institutions can more accurately identify borrowers at risk of default. The first approach is to use the K-Means-SMOTE algorithm to eliminate the imbalance within the class. In the second step, ResNet is used for feature extraction, and then two-layer LSTM is used for learning to strengthen the ability of neural networks to mine and utilize deep information. Finally, the model performance is improved by using the IDWPSO algorithm for optimization when debugging the neural network.

On two unbalanced datasets (category ratios of 700:1 and 3:1 respectively), the multi-stage improved model was compared with ten other models using accuracy, precision, specificity, recall, G-measure, F-measure and the nonparametric Wilcoxon test. It was demonstrated that the multi-stage improved model showed a more significant advantage in evaluating the imbalanced credit dataset.

In this paper, the parameters of the ResNet-LSTM hybrid neural network, which can fully mine and utilize the deep information, are tuned by an innovative intelligent optimization algorithm to strengthen the classification performance of the model.

The textile industry holds immense significance in the economy of any nation, particularly in the production of synthetic yarn and fabrics. Consequently, the pursuit of acquiring high-quality products at a reduced cost has become a significant concern for countries. The primary objective of this research is to leverage data mining and data intelligence techniques to enhance and refine the production process of texturized yarn by developing an intelligent operating guide that enables the adjustment of production process parameters in the texturized yarn manufacturing process, based on the specifications of raw materials.

This research undertook a systematic literature review to explore the various factors that influence yarn quality. Data mining techniques, including deep learning, K-nearest neighbor (KNN), decision tree, Naïve Bayes, support vector machine and VOTE, were employed to identify the most crucial factors. Subsequently, an executive and dynamic guide was developed utilizing data intelligence tools such as Power BI (Business Intelligence). The proposed model was then applied to the production process of a textile company in Iran 2020 to 2021.

The results of this research highlight that the production process parameters exert a more significant influence on texturized yarn quality than the characteristics of raw materials. The executive production guide was designed by selecting the optimal combination of production process parameters, namely draw ratio, D/Y and primary temperature, with the incorporation of limiting indexes derived from the raw material characteristics to predict tenacity and elongation.

This paper contributes by introducing a novel method for creating a dynamic guide. An intelligent and dynamic guide for tenacity and elongation in texturized yarn production was proposed, boasting an approximate accuracy rate of 80%. This developed guide is dynamic and seamlessly integrated with the production database. It undergoes regular updates every three months, incorporating the selected features of the process and raw materials, their respective thresholds, and the predicted levels of elongation and tenacity.

This chapter focuses on a case study of Kareem Exchange Company and its use of big data analysis to detect and prevent fraud and suspicious financial transactions. The chapter describes the various phases of the big data analysis cycle, including discovery, data preparation, model planning, model building, operationalization, and communicating results, and how the Kareem Exchange Company team implemented each phase. This chapter emphasizes the importance of identifying the business problem, understanding the resources and stakeholders involved, and developing an initial hypothesis to guide the analysis. The case study results demonstrate the potential of big data analysis to improve fraud detection capabilities in financial institutions, leading to informed decision making and action.

As convenience and anonymity, people with mental illness are increasingly willing to communicate and share information through social media platforms to receive emotional and spiritual support. The purpose of this paper is to identify the degree of depression based on people's behavioral patterns and discussion content on the Internet.

Based on the previous studies on depression, the severity of depression is divided into four categories: no significant depressive symptoms, mild MDD, moderate MDD and severe MDD, and defined each of them. Next, in order to automatically identify the severity, the authors proposed social media digital cues to identify the severity of depression, which include textual lexical features, depressive language features and social behavioral features. Finally, the authors evaluate a system that is developed based on social media digital cues in the experiment using social media data.

The social media digital cues including textual lexical features, depressive language features and social behavioral features (F1, F2 and F3) is the relatively best one to classify four different levels of depression.

This paper innovatively proposes a social media data-based framework (SMDF) to identify and predict different degrees of depression through social media digital cues and evaluates the accuracy of the detection through social media data, providing useful attempts for the identification and intervention of depression.

This study aims to compare machine learning models, datasets and splitting training-testing using data mining methods to detect financial statement fraud.

This study uses a quantitative approach from secondary data on the financial reports of companies listed on the Indonesia Stock Exchange in the last ten years, from 2010 to 2019. Research variables use financial and non-financial variables. Indicators of financial statement fraud are determined based on notes or sanctions from regulators and financial statement restatements with special supervision.

The findings show that the Extremely Randomized Trees (ERT) model performs better than other machine learning models. The best original-sampling dataset compared to other dataset treatments. Training testing splitting 80:10 is the best compared to other training-testing splitting treatments. So the ERT model with an original-sampling dataset and 80:10 training-testing splitting are the most appropriate for detecting future financial statement fraud.

This study can be used by regulators, investors, stakeholders and financial crime experts to add insight into better methods of detecting financial statement fraud.

This study proposes a machine learning model that has not been discussed in previous studies and performs comparisons to obtain the best financial statement fraud detection results. Practitioners and academics can use findings for further research development.