Journal of Advanced Research Design

Application of Genetic Algorithm in Training Automatic Speech Recognition

2025-02-27T09:50:14+07:00

The application of a genetic algorithm in training automatic speech recognition is considered. In the learning process, hidden Markov models are used to evaluate the statistical properties of each word, including the sequence of cepstral coefficients, as well as the transition probabilities between model states. As the model trains, it seeks to improve the fit between the input cepstral coefficients and the target words in order to improve the accuracy of speech recognition. Once trained, the system is used to recognize new speech inputs and identify the corresponding words. This is done by applying trained Hidden Markov Models to new sequences of cepstral coefficients. Genetic algorithms can be useful for optimizing some aspects of speech recognition systems, such as selecting optimal parameters for processing speech signals and choosing the most appropriate models for specific tasks. The results of a comparative analysis based on typical examples are presented, demonstrating the advantages and high efficiency of the genetic algorithm. The main advantages of the GA are indicated, such as independence from initial conditions and the ability to overcome local extrema, which makes it a promising tool for training speech recognition models. The conclusion emphasizes the significance of the research results and their contribution to the development of SMM training methods for speech recognition.

Assessment of Time Series Model for Predicting Long-Interval Consecutive Missing Values in Air Quality Dataset

2025-02-27T10:05:50+07:00

Air pollutant concentration in Malaysia is continuously monitored using the Continuous Ambient Air Quality Machine (CAAQM). During the observation phase by CAAQM, some air pollutant datasets were detected missing due to machine failure, maintenance, position changes and human error. Incomplete datasets especially with the longer gaps of consecutive missing observation may lead to several significant problems including loss of efficiency, difficulties in using some computational software and bias estimation due to differences of observed and predicted dataset. This study aim evaluates the performance of the time series method i.e. Auto Regression Integrated Moving Average (ARIMA) for filling long hours of missing data in an air pollution dataset. The dataset of PM10, SO2, NO2, O3, CO, wind speed, relative humidity and ambient temperature for Pegoh and Kota Kinabalu in 2018 were used for analysis. Monte Carlo Markov Chain (MCMC) and Expectation-Maximization (EM) were employed to compare with ARIMA's effectiveness in filling the simulated missing gaps in air quality dataset. Existing missing data in the raw data were pre-treated and then simulated into 5%, 10% and 15% of missing data ranging from 24-hour to 120-hour intervals. The performance of the imputation approach was assessed using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Prediction Accuracy (PA) and Index of Agreement (IA). Overall, the Expectation-Maximization technique was selected the most effective at filling in simulated long gaps of missing data of air pollutant dataset with the range of IA from 0.74 to 0.77. In contrast, the ARIMA approach performed poorly in this research with range of IA value of 0.44 to 0.48. This was because of it requires past time-series data to generalize a forecast or impute missing data, hence, the forecast becomes a straight line and performed poorly at predicting series with long hours of missing observation.

Investigating the Determinants of Cloud Computing-Software as a Service Adoption in Pakistani SMEs from the Perspective of SME Managers

2025-02-27T10:33:09+07:00

Pakistan's small and medium enterprises (SMEs) sector oppose many challenges, encompassing resource mismanagement, infrastructural deficiencies, a burgeoning volume of data and impediments hindering organizational growth. These challenges include financial constraints from corruption and insecurity and limited IT resources and infrastructure access. Consequently, adopting Cloud Computing Software as a Service (CC-SaaS) presents a potentially advantageous solution. It offers SMEs an avenue to enhance operational efficiency through cost-effective technological integration, facilitating improved e-services for citizens and promoting knowledge sharing for increased benefits. However, the adoption of CC-SaaS remains relatively limited within the Pakistani SME landscape, beset by numerous obstacles, like data privacy, legal compliance and security concerns. Despite the extensive literature on factors influencing cloud computing (CC) adoption, most of these studies emanate from developed nations. More attention has been devoted to examining the adoption of CC-SaaS. Indeed, within countries marred by conflict and instability, the adoption landscape of CC-SaaS remains largely uncharted. This study addresses this knowledge gap by investigating the determinants influencing organizational intentions regarding adopting CC-SaaS among Pakistani SMEs operating in an environment fraught with conflict. It necessitates examining how the prevailing conflict dynamics in Pakistan might impact CC-SaaS adoption within the country. Ultimately, this research aspires to benefit SME organizations in Pakistan and contribute to validating measurement frameworks for future studies. The study's conceptual model draws upon the Technology-Organization-Environment theory and the Organization Support Theory. Expert opinions informed the identification of moderating effects. Data was collected from a sample of 368 SME managers operating in Pakistan. Subsequently, the collected data underwent analysis employing the Partial Least Squares Structural Equation Modeling (PLS-SEM) approach. The study findings indicate that the impacts of technology organization, environment and Organization Support Theory on CC-SaaS adoption are statistically significant.

Decision Analysis on the Customer Preferences of Automobiles using AHP-TOPSIS Model

2025-02-14T09:33:58+07:00

Vehicle registration has surpassed the number of populations in Malaysia in 2021. Automobiles remain the preferred mode of transport with the top brands including Perodua, Proton, and Toyota in Malaysia. The local automotive industry is facing various challenges such as robotics and automation, sustainability, and resource scarcity in the supply chain. Therefore, this paper aims to propose a research framework for studying the consumer preferences of local car models in Malaysia using the integration of Analytic Hierarchy Process (AHP) and the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) models. In the first stage, AHP model is used to determine the weights of main decision criteria and sub-criteria. In the second stage, the weights of the decision alternatives are determined using TOPSIS model. In this study, the main decision criteria include design, convenience, performance, safety and economic aspect. This study shows that safety, economic aspect, and design are the top three main decision criteria in the consumer preferences of local car models. In addition, car price, engine capacity, style and aerodynamic designs are the three most influential sub-criteria affecting consumer preferences of local car models. Besides, Perodua Myvi appears as the most preferred car model based on the highest relative closeness coefficient, followed by Proton Persona, Perodua Bezza, Proton Saga, Perodua Axia and Proton Iriz. This paper contributes by providing insights to local automotive industry to improve based on the top influential criteria affecting car model selection in Malaysia.

Optimization of Artificial Neural Networks using Genetic Algorithm for Palm Fruit Ripeness Classification

2025-02-26T13:57:16+07:00

The cultivation of oil palm has become a major contributor to vegetable oil production, playing a crucial role in meeting the global demand for food in the agricultural industry. The efficiency of palm oil extraction heavily relies on accurate fruit ripeness classification, which is traditionally performed manually. Accurate assessments of oil palm fruit maturity levels before processing are crucial for Malaysian producers and exporters. This report outlines a machine-learning approach for sorting palm fruit by utilizing an artificial neural network (ANN) and optimizing its performance via a genetic algorithm (GA). The research specifies input features such as colour, size, texture and output classes that include the phases of unripe, ripe and overripe palm fruits. The Palm Fruit Database was built by acquiring images of palm fruits through the on-site collection and an e-database called ROBOFLOW and then defining their maturity characteristics through image processing techniques. Using MATLAB software, this study applied three different ANN training methods: the Levenberg-Marquardt backpropagation algorithm, the Bayesian regularization backpropagation algorithm and the scaled conjugate gradient backpropagation algorithm, resulting in accuracy scores of 74.96, 86.19 and 74.97% respectively. The Bayesian regulation has the best performance. The initial accuracy of the ANN test was 77%, but after applying GA optimization, the accuracy improved to 87%, a 10% increase. The proposed approach offers a practical solution for improving the efficiency and accuracy of palm fruit classification in the palm oil industry, ultimately leading to better quality control and increased productivity.

AI Innovation in Architectural Design: Enhancing Aesthetic Experience with 'Midjourney'

2025-02-03T16:06:33+07:00

This study deeply explores the application of artificial intelligence (AI) technology represented by Midjourney in the field of modern architectural design and its influence on architectural aesthetics. With the progressive improvement of AI innovation in a few areas, particularly in architectural design, it has risen above the confinements of conventional design strategies and opened up modern measurements for the expression of architectural aesthetics. Midjourney technology not only promotes design innovation but also improves the efficiency and flexibility of the design process. Compared to conventional architectural design software such as AutoCAD, Revit and SketchUp, Midjourney shows significant advantages in innovation and user experience. In addition, the study discusses the potential of AI technology to drive the development of architectural aesthetics, especially in generating architectural renderings through AI, which presents unique visual and sensory experiences. Even though the application of AI innovation in architectural design still faces numerous challenges, such as not only technical adaptability, and cost-effectiveness issues, but also application in several social and natural settings. This study points out the direction for future research. By and large, this consideration gives unused viewpoints for understanding the part of AI innovations in present architectural design and offers important references and bits of knowledge for designers, creators, and related experts.

Recent Progress in Proppant Technology for Improving the Fracture Conductivity in Hydraulic Fracturing

2025-02-14T10:52:35+07:00

Proppants play a crucial role in hydraulic fracturing (HF) operations and in sustaining conductive fractures during well production. However, challenges persist regarding their resilience to closure stress and downhole conditions. Coatings have emerged as a promising solution to enhance proppant efficacy, particularly in addressing mechanical failure. This study highlights the recent advancement in proppant technology and focuses specifically on the impact of different resins coated proppants in improving the fracture conductivity after HF operation. Polymer coatings, especially thermosetting-based resin coatings are widely used due to their ability to improve both the strength and flexibility of the coated proppants. Proppants coated with a thin layer of resin offer several advantages, including good permeability, shape improvement and lower cost compared to regular coatings. Additionally, the incorporation of nanomaterials into resin coatings has shown promising results in augmenting proppant durability and flow conductivity as well as enhancing embedment prevention, reducing the generated fines after proppant crushing and improving the overall oil and gas production rate. For that, a summary was presented of the latest academic discussions and conclusions on the impact of resin coating on proppant and its pivotal role in enhancing proppant performance which led to increased fracture conductivity. Leveraging insights garnered from these discussions can positively contribute to the sustainable extraction of hydrocarbon resources in the oil and gas industry.

GF-CNN: A Hybrid Approach for Pollen Recognition Combining Gabor Filters and Convolutional Neural Networks

2025-03-03T14:49:42+07:00

Pollen identification is a critical task across various scientific disciplines, including geology, ecology, evolutionary biology and botany. However, existing methods for pollen identification are often labour-intensive, time-consuming and dependent on highly skilled experts, highlighting the need for an automated and precise system. This study introduces an innovative approach that combines Gabor Filters (GF) with Convolutional Neural Networks (CNN) to enhance the accuracy of pollen classification. The Gabor filters are applied to high-resolution images of diverse pollen species, accentuating texture-specific details essential for differentiation. These pre-processed images are subsequently analysed using a CNN architecture with multiple layers designed to discern hierarchical features critical for precise classification. The proposed GF-CNN model demonstrates exceptional proficiency, achieving remarkable accuracy rates of 99.85% for the Malaysian Pollen Dataset (MPD) and 99.43% for the New Zealand Pollen Dataset (NPD). These results underscore the model's ability to balance precision and recall effectively. Additionally, the model exhibits high sensitivity, indicating an increased true-positive rate, which is essential for detailed ecological studies. Furthermore, the model's improved specificity scores highlight its success in minimizing false positives, emphasizing its relevance for precision-focused research.

Physical Characterization of Non-Woven Treated and Untreated Palm Empty Fruit Bunch Layered by Fibreglass as an Interior Insulation Material

2025-01-31T14:51:45+07:00

Palm empty fruit bunch (PEFB) fibres hold potential as a natural material for insulation, offering an alternative to the commonly used fibreglass in HVAC system ducting. Fibreglass (FG) in commercial insulation materials contributes to issues such as unpleasant environmental noise, which can have harmful effects on communication, health and overall quality of life. This study aims to prepare nonwoven untreated and treated PEFB with FG at different ratios using a silane coupling agent treatment. The nonwoven materials are produced through a combination of a needle punching machine and hot compress technique and various characteristics, including thermal decomposition, functional groups, thermal conductivity and water contact angles, are examined.

An Evaluation of EVM-Compatible Blockchain Platforms for Trade Finance

2025-02-13T15:52:01+07:00

Blockchain, such as Bitcoin and Ethereum, has received significant attention and widespread usage in recent years. However, blockchain scalability has emerged as a challenging issue. This article explores the existing scalability options for blockchain, which can be categorized into two groups: first layer solutions and second layer solutions. First layer solutions involve network modifications like altering block size, while second layer solutions encompass techniques applied outside of the blockchain. Ethereum, the second largest blockchain, utilizes the Ethereum Virtual Machine (EVM) for executing smart contracts on the blockchain. Currently, there are several EVM-compatible blockchains with noticeable differences. In this study, we evaluated multiple platforms for conducting business processes in trade finance. We considered both Layer 1 and Layer 2 blockchain solutions and examined variations in cost and performance (speed). Based on the evidence gathered in this study, we provide recommendations for system designers to consider when selecting a blockchain platform.

Challenges of Blockchain Technology and its Relationships to Sustainable Education: An Analysis using AI-Based Literature Review

2025-02-25T10:00:29+07:00

Blockchain technology is known for its features like immutability, transparency, decentralisation and security, which can enhance trust, create transparency and promote accountability in educational systems. This study explores the relationship between blockchain technology and sustainable education, focusing on the challenges and benefits of integrating blockchain into educational systems. The research employs an AI-based literature review using Scite.AI to gather and analyse relevant articles, ensuring the selected literature directly addresses blockchain technology within the context of sustainable education. The study identifies significant challenges, such as organisational, technological and environmental factors, that must be addressed to leverage blockchain for sustainable educational advancements. The research highlights that integrating blockchain in education can support sustainable practices by ensuring secure credentialing, efficient resource allocation and fostering trust in digital learning environments. Additionally, blockchain technology can ease worldwide accreditation systems for teaching, learning, practice and business communication, ensuring data protection and transmission of student projects and evaluations. The study also points out the need for robust frameworks to address blockchain technology's technical, ethical and policy challenges in education. It proposes a new Relationship model between Blockchain technology and Sustainable Education. The findings suggest that while blockchain technology has the potential to revolutionise education by providing creative and environmentally responsible educational tools, its implementation faces several technological organisational and environmental challenges. Future research directions are suggested to address these gaps and challenges, aiming to develop effective strategies for integrating blockchain technology into sustainable education systems.