Conventional Deep Neural Network (DNN) learning process built up the DNN architecture to fit the input space, in terms of size, structure and neural shapes. But the calibration for the input data space characterisation is never realised. This paper reports “Less is More” principle and its implementation in the development of DNN. The recent phenomenal success of artificial DNN has brought a popular but false belief: the deeper the network structure and the more training data, the better. We believe “Less is More”, as the characteristic of a problem is more fundamental than the association of statistics only. We view each input sample as a composition of linked pixels in a unique pattern, and the features of such linkage should be characterised. Technically, the necessary feature for each sample is embedded in a pixel-based graph and can be extracted, typically ~50 features per sample pattern. With these 50 characteristic features, we extracted a neater network (size of input, hidden and output layers are 64, 6 and 7 neurons, respectively), and achieved the highest accuracy (~99.9% against all the published DNNs regardless how deep, physical or digital they are) and a higher computing efficiency (~58% less CPU runtime). The developed DNNs bears the features of circuitry which offers better explainability. The successful application of Less is More approach supports and promotes its wider applications; it contributes to DNN methodology. We also anticipate that the developed methodology will eliminate the waste of unnecessary computational efforts such as pruning et al.

Assessment is essential to educational system. Automatic grading reduces the time and effort taken by tutors to assess the answers written by the students. To understand recent computational methods used for automatic grading, a review has been conducted. 4,084 articles were initially identified using a keyword search. After filtering, the number was reduced to 57. It was found that statistical models are normally used in Automatic-Short-Answer-Grading (ASAG); vector-based similarity measures are the most popular among projects; pilot datasets are mostly used; standard datasets for evaluation are missing. Evidence shows that machine learning and deep learning are most popularly adopted methods and generative AI, e.g., LLMs and ChatGPT are also jump to the chance, which indicates that integrating AI in education is an inevitable trend. Also, most investigations prefer to adopt multiple approaches to improve computational quality, dataset analysis, and evaluation results. The identified research gaps will be a useful reference guide to users/researchers beneficial to formative/summative assessment. We concluded that the presented outcome, analysis and discussions are informative to academia and pedagogical practitioners who are interested in further developing/using ASAG systems. Although research into ASAG is still rudimentary, it is a promising area with impact on academic circles/commercially educational markets.

Autism Spectrum Disorder (ASD) presents challenges in early screening due to its varied nature and sophisticated early signs. From a machine-learning (ML) perspective, the primary challenges include the need for large, diverse datasets, managing the variability in ASD symptoms, providing easy-to-understand models, and ensuring ASD predictive models that can be employed across different populations. Interpretable or explainable classification algorithms, like rule-based or decision tree, play a crucial role in dealing with some of these issues by offering classification models that can be exploited by clinicians. These models offer transparency in decision-making, allowing clinicians to understand reasons behind diagnostic decisions, which is critical for trust and adoption in medical settings. In addition, interpretable classification algorithms facilitate the identification of important behavioural features and patterns associated with ASD, enabling more accurate and explainable diagnoses. However, there is a scarcity of review papers focusing on interpretable classifiers for ASD detection from a behavioural perspective. Thereby this research aimed to conduct a recent review on rule-based classification research works in order to provide added value by consolidating current research, identifying gaps, and guiding future studies. Our research would enhance the understanding of these techniques, based on data used to generate models and obtain performance by trying to highlight early detection and intervention ways for ASD. Integrating advanced AI methods like deep learning with rule-based classifiers can improve model interpretability, exploration, and accuracy in ASD-detection applications. While this hybrid approach has feature selection relevant features that can be detected in an efficient manner, rule-based classifiers can provide clinicians with transparent explanations for model decisions. This hybrid approach is critical in clinical applications like ASD, where model content is as crucial as achieving high classification accuracy.

Alzheimer’s disease (AD) poses significant challenges for the elderly, leading to cognitive decline, social isolation, and lower quality of life. Current interventions often require cumbersome wearable devices e.g. the camera-based monitoring that may raise privacy concerns. However, these issues are not fully addressed previously. To fill this gap, this research proposes a new framework in non-invasive combination of Virtual Reality (VR), Voice recognition, and Artificial Intelligence (AI) to act as a supportive system for people with AD. The system provides a brand-new approach that tailored cognitive stimulation and companionship through the immersive VR scenarios, memory games, virtual trips, and an AI assistant together as a single platform. The AI-based assessment of the patient is employed to ensure that the experience is more relevant and helpful to the patient. The voice recognition is the most simple and easy user-interface. The security measures include access controls, encryption and continuous monitoring of cloud patient data. The initial study has been promising evidenced by the outcome of involving patients with Alzheimer’s and dementia, their families, and clinicians. Participants reported heightened interest, better quality of life, less sense of isolation, and improved cognitive functioning, which have particularly achieved one of our goals, in patients’ well-beings in mental healthcare. The research indicates a significant step forward enhancing the quality of support for both cognitive function and social interaction for older adults with AD and dementia. In comparison with other currently existing systems, our newly developed integrated framework has made additional contributions to the areas of AD in dynamic cognitive adaptation, bilingual interaction, and secure real-time personalized system.

The transportation sector significantly contributes to global greenhouse gas emissions, particularly CO2. This research aims to develop a user-friendly web application that accurately predicts CO2 emissions of light-duty vehicles using machine learning models. By using datasets from the UK Vehicle Certification Agency (VCA), various regression models such as Decision Tree, Random Forest, and Gradient Boosting, were trained and evaluated for predicting CO2 emission.The model was deployed as a Streamlit web application to allow users to estimate vehicle emissions based on input variables. This research underscores the potential of machine learning and artificial intelligence in supporting the digital transformation of the transportation sector, providing a cost-effective tool for stakeholders to assess CO2 emissions and support environmental sustainability. Future work involves exploring additional machine learning algorithms and enhancing the web application for broader use.

Artificial Intelligence (AI) and Large Language Models (LLMs) significantly have changed educational paradigms, enabling a significant reinvention in how teaching and learning happens today. Through this systematic literature review, the paper aims to offer a deep understanding of how AI and LLMs are implemented in educational settings today, providing examples of where these tools were used for personalization (both targeting students’ academic performance and choosing content that aligns best with learner’s aspirations), accessibility enhancements as well as their potential applications related to resource management automation. The central themes probed include the application of AI to help deploy appropriate learning strategies according to individual learner profiles. LLMs for improved responsive and interactive education tools, and the potential for autonomous assessment machines with associated feedback mechanisms.The opportunities for the aid of AI and LLMs to improve teaching and learning are vast, but this review will also question some associated challenges such as ethical issues[1], data privacy concerns and risks of algorithmic bias[2]. It is concluded by suggesting future research and application directions for the ethical development of inclusive technologies, which are secure from a privacy point of view too. This is followed by a reflection on the strategic implications of such technologies for those in charge of education policy and development hoping ideas related to how AI-LLMs rejoice more efficacious learning experiences could be born out.

Low-code development has gained significant recognition in industry and academia. However, lack of reusability is inherent to existing low-code development platforms. Based on the literature review and practical evaluation, this paper highlights the importance of platform extensibility. Modern technologies are reviewed, among which the essential components are selected for an extensible and open-source low-code development platform. The challenges of designing platform architecture and developing synchronized editable textual and visual representations are ready for us to take in future directions.

The minimum creep rate and the Monkman-Grant relationship are currently essential methods for predicting creep life. The accuracy of traditional power law minimum creep rate equation is unsatisfactory when applied to a wide range of stresses and temperatures. In this study, four sets of Grade 91 minimum creep rate experimental data on different stress and temperatures were applied for power law, hyperbolic sine (HS) law and a modified hyperbolic sine (MHS) law minimum creep rate equation. Additionally, attempts have been made to introduce the effects of temperature into the modified hyperbolic sine law equation. The result shows that the modified hyperbolic sine law equation exhibits higher accuracy to a wide range of stress and temperature level and shows a good agreement with life prediction. The extrapolation curve from high stress to low stress matches the trend of experimental data. It is more suitable for extrapolation to long-term service facilities such as power plants with a design life of up to 60 years.

The creep rupture model developed based on creep cavitation is a physically based method for predicting creep life. By combining creep damage mechanisms with experimental data to establish a creep rupture criterion, this approach provides greater confidence in applications across a wide range of stress level (especially low stress long-term creep) and for extrapolation. This study reported the creep lifetime prediction of a creep rupture model based on creep cavitation for Grade 91 steel over a wide range of stress level. A temperature-dependent function was introduced to creep rupture model to assess the accuracy and reliability across muti-temperatures. The results show that the creep rupture model developed based on creep cavitation achieves high accuracy over a wide range of stress level (1MPa to 450MPa) and temperature. The trend of lifetime prediction curves has a good agreement with experimental data, shows great extrapolation potential. More reliable for long-term low-stress creep life prediction.

Climate change caused by greenhouse gas (GHG) emissions is an escalating global issue, with the transportation sector being a significant contributor, accounting for approximately a quarter of all energy-related GHG emissions. In the transportation sector, vehicle emissions testing is a key part of ensuring compliance with environmental regulations. The Vehicle Certification Agency (VCA) of the UK plays a pivotal role in certifying vehicles for compliance with emissions and safety standards. One of the primary methods employed by the VCA to measure vehicle emissions for light-duty vehicles is the Worldwide Harmonized Light Vehicles Test Procedure (WLTP). The WLTP is a global standard for testing vehicle emissions and fuel consumption, and sensors are crucial in ensuring accurate, real-time data collection in laboratories. Using the data collected by the VCA, regression machine learning models were trained to predict CO2 emissions in light-duty vehicles. Among six regression models tested, the Decision Tree Regression model achieved the highest accuracy, with a Mean Absolute Error (MAE) of 2.20 and a Mean Absolute Percentage Error (MAPE) of 1.69%. It was then deployed as a web application that provides users with accurate CO2 emission estimates for vehicles, enabling informed decisions to reduce GHG emissions. This research demonstrates the efficacy of machine learning and AI-driven approaches in fostering sustainability within the transportation sector.

In high-temperature applications such as power generation and aerospace, understanding material failure due to creep deformation is crucial, especially for components like jet engine compressor blades and steam turbines. This study investigates the evolution of creep cavitation damage, focusing on cavity nucleation, growth, and coalescence in the material's microstructure. Beyond nucleation and growth, this study investigates cavity coalescence, an inevitable process leading to microcracks and macrocracks. It captures the evolution of exact cavity measurements. This mechanism is associated with an increasing nucleation rate and a decreasing growth rate of individual cavities. Enhancing the accuracy of tensile creep behaviour models, this study extends the cavitation model initially proposed by Riedel and calibrated by Qiang Xu. It addresses cavity coalescence during late-stage rupture using high-fidelity experimental data and a modelling approach. A spherical reconstruction hypothesis for cavity coalescence is proposed and validated with cavitation data from the brass alloy Cu-40Zn-2Pb. The model demonstrates accuracy for lifetime modelling with an increase of accuracy from traditionally 45–83% and a coefficient of determination from 0.767 to 0.997, highlighting its robustness for describing coalescence and aiding in design and maintenance assessments.

This paper is to discuss a recent emerging research area in AI related application in the domain of engineering education. The background study demonstrates the importance and significance of the research to be conducted. The research questions are proposed. Methods and expected outcome are discussed. Finally, the paper is ended with the possible benefits and impact on the application domain in near future.

Colon cancer is a significant global health problem, and early detection is critical for improving survival rates. Traditional detection methods, such as colonoscopies, can be invasive and uncomfortable for patients. Machine Learning (ML) algorithms have emerged as a promising approach for non-invasive colon cancer classification using genetic data or patient demographics and medical history. One approach is to use ML to analyse genetic data, or patient demographics and medical history, to predict the likelihood of colon cancer. However, due to the challenges imposed by variable gene expression and the high dimensionality of cancer-related datasets, traditional transductive ML applications have limited accuracy and risk overfitting. In this paper, we propose a new hybrid feature selection model called HMLFSM–Hybrid Machine Learning Feature Selection Model to improve colon cancer gene classification. We developed a multifilter hybrid model including a two-phase feature selection approach, combining Information Gain (IG) and Genetic Algorithms (GA), and minimum Redundancy Maximum Relevance (mRMR) coupling with Particle Swarm Optimization (PSO). We critically tested our model on three colon cancer genetic datasets and found that the new framework outperformed other models with significant accuracy improvements (95%, ~97%, and ~94% accuracies for datasets 1, 2, and 3 respectively). The results show that our approach improves the classification accuracy of colon cancer detection by highlighting important and relevant genes, eliminating irrelevant ones, and revealing the genes that have a direct influence on the classification process. For colon cancer gene analysis, and along with our experiments and literature review, we found that selective input feature extraction prior to feature selection is essential for improving predictive performance.

Autism Spectrum Disorder (ASD) is a significant healthcare concern due to the large number of cases detected annually, and the massive resources required to support individuals on the spectrum and their families. Data mining and artificial intelligence (AI) techniques have shown promising results in research on healthcare applications, including ASD diagnosis, by providing accurate diagnosis. However, most data models developed by these intelligent techniques, a) do not provide details behind the diagnostic decision to the stakeholders such as clinicians, patients, and caregivers, and b) are criticised for being biased to a single data model rather a group of models. A model that can interpret results involved in the diagnostic process is advantageous offering digital knowledge to healthcare professionals besides adhering to the General Data Protection Regulation (GDPR) terms primarily ‘results derived by automated decision-making methods' like AI techniques. More essentially, when the prediction is performed by a group of models this can reduce the decision bias of the diagnosis. This article fills these gaps by proposing a framework based on ensemble learning where a rule-based classifier develops interpretable data models for ASD diagnosis.

Clustering is one of the challenging machine learning techniques due to its unsupervised learning nature. While many clustering algorithms constrain objects to single clusters, K-means overlapping partitioning clustering methods assign objects to multiple clusters by relaxing the constraints and allowing objects to belong to more than one cluster to better fit hidden structures in the data. However, when datasets contain outliers, they can significantly influence the mean distance of the data objects to their respective clusters, which is a drawback. Therefore, most researchers address this problem by simply removing the outliers. This can be problematic especially in applications such as fraud detection or cybersecurity attacks risk analysis. In this study, an alternative solution to this problem is proposed that captures outliers and stores them on-the-fly within a new cluster, instead of discarding. The new algorithm is named Outlier-based Multi-Cluster Overlapping K-Means Extension (OMCOKE). Empirical results on real-life multi-label datasets were derived to compare OMCOKE’s performance with other common overlapping clustering techniques. The results show that OMCOKE produced a better precision rate compared to the considered clustering algorithms. This method can benefit various stakeholders as these outliers could have real-life applications in cybersecurity, fraud detection, and the anti-phishing of websites.

Industrial control system security is essential to protecting both industrial output and Critical National Infrastructure (CNI) from cyber-attack, as well as ensuring the safety of workers, members of the public and the environment.Operators of these facilities face a security problem in the dearth of effective countermeasures. This paper addresses this problem by identifying the reasons why so few countermeasures exist and what approaches could be taken to remedy this in a manner that serves both traditional Industrial Control Systems (ICSs), and the emerging needs of the Industrial Internet of Things (IIoT).Circa two-thousand-five-hundred documents (sourced from a combination of five academic search engines, standards agencies, and industrial reports) were reviewed and analysed. From this was found that existing ICS countermeasures are largely derived from existing IT solutions that do not seek to take advantage of the specific characteristics of ICSs – making them less effective or inappropriate in many ICS applications; this is particularly true of network intrusion protection systems, for which false positive detection can have a serious impact on the safe and reliable operation of industrial facilities.It is proposed that the characteristics of ICS and IIoT communications networks lend themselves to a whitelisting approach to network intrusion protection, which would avoid the problem of false positives, and that future work based on the OPC-UA protocol would prove this and demonstrate its suitability for all ICS and IIoT applications.

The application of Artificial Intelligence or AI in education has been the subject of academic research for more than 30 years. The field examines learning wherever it occurs, in traditional classrooms or at workplaces so to support formal education and lifelong learning. It combines interdisciplinary AI and learning sciences (such as education, psychology, neuroscience, linguistics, sociology and anthropology) in order to facilitate the development of effective adaptive learning environments and various flexible, inclusive tools. Nowadays, there are several new challenges in the field of education technology in the era of smart phones, tablets, cloud computing, Big Data, etc., whose current research questions focus on concepts such as ICT-enabled personalized learning, mobile learning, educational games, collaborative learning on social media, MOOCs, augmented reality application in education and so on. Therefore, to meet these new challenges in education, several fields of research using AI have emerged over time to improve teaching and learning using digital technologies. Moreover, each field of research is distinguished by its own vision and methodologies. In this article, to the authors present a state of the art finding in the fields of research of Artificial Intelligence in Education or AIED, Educational Data Mining or EDM and Learning Analytics or LA. We discuss their historical elements, definition attempts, objectives, adopted methodologies, application examples and challenges.

ELD, which stands for Economic Load Dispatch is a time uncontrollable and hard problem in electrical engineering. It is concerned with minimizing the cost or price of economic manufacture, accordingly, assigning the power produced through an individual unit in the utmost economic way possible. It is essential for methods, which don’t impose constraints on the form of the fuel cost curvatures due to the nonlinear features of the units. Classical calculus-based techniques are incapable of satisfactorily addressing these types of problems. Alternatively, metaheuristic algorithms are very common nowadays. However, due to their stochastic nature, they tend to perform poorly when dealing with large size ELD problems. Therefore, in this research work, an innovative amalgam metaheuristic optimization algorithm is presented to tackle the above issue. The algorithm is composed of the Levy Flight method and two well-known metaheuristic algorithms, which are Particle Swarm Optimization Algorithm (PSO) and Salp Swarm Algorithm (SSA). The suggested algorithm is evaluated against five robust, competitive, and modern metaheuristic algorithms, which are found in the literature. The results show that the proposed algorithm outpaced all the other competitive algorithms in terms of effectiveness and efficiency.