Yi-Fei Tan

Wearable technology has become an element of health tracking and occupies a substantial place in growing the Internet of Medical Things (IoMT). IoMT can minimize the mortality rate due to the early diagnosis of the diseases. Although this area is improving, there is still a way to increase the accuracy of prediction. To resolve this, the paper proposes Soft computing enforced Adaptive Neuro-Fuzzy Inference System (ANFIS) model for heart disease prediction, designed to ensure reliable diagnosis. The model integrates fuzzy logic and neural networks, optimized by Particle Swarm Optimization (PSO), to address challenges such as noisy data and local minima in training. Additionally, privacy-preserving mechanisms are incorporated, ensuring optimal data flow and compliance with healthcare regulations: Feature Extraction using Particle Swarm Optimization (PSO) helps in finding the pertinent features of clinical data and also finds the best feature space of clinical data so that the most promising features can be selected to make reliable predictions. Adaptive Neuro-Fuzzy Inference System (ANFIS), which can be used to classify heart disease data (and apply neural network and fuzzy logic), suffers problems with local minima because of the gradient nature of this learning. The hybrid ANFIS-PSO model enhances diagnostic accuracy by improving feature extraction and parameter optimization, yielding higher performance than traditional models. The hybridisation of PSO and ANFIS allows attaining higher predictive accuracy and system reliability, highlighting the significance of the optimisation of the feature extraction and parameters setting to gain optimum performance of the health monitoring application.

The advent of Artificial Intelligence (AI) has revolutionized the financial industry by enabling more accurate, efficient, and dynamic decision-making processes. This paper explores the transformative role of AI in financial risk assessment and the development of predictive tools that facilitate smarter financial decisions. It investigates how machine learning algorithms, natural language processing, and neural networks are deployed to assess credit risk, forecast market trends, detect fraud, and enhance portfolio management. By synthesizing recent advancements and real-world applications, this study evaluates the efficacy, reliability, and ethical considerations of AI-driven tools in finance. The paper also addresses the challenges of data quality, algorithmic bias, and regulatory compliance. Through a comprehensive analysis, it provides insights into the current landscape and future prospects of AI in shaping a resilient and intelligent financial ecosystem.

Modern healthcare depends much on personalized treatment optimization, which seeks to improve patient outcomes by customizing medical procedures depending on particular health circumstances. This work presents a hybrid fuzzy-deep learning model (HF-DLM) to maximize treatment plans in smart healthcare systems. Combining fuzzy logic with deep learning, the approach uses deep neural networks for pattern identification and decision-making to manage ambiguity in medical data: While deep learning increases prediction accuracy by automatic feature extraction, the fuzzy component improves interpretability by including expert knowledge. Clinical datasets and actual electronic health records (EHRs) help to assess the proposed HF-DLM. HF-DLM beats traditional machine learning and rule-based systems in forecasting ideal treatment regimens, thereby lowering side effects, and so enhancing patient recovery rates. Comparative study of current methods emphasizes in terms of accuracy, recall, and computing efficiency the benefits of HF-DLM. The paper also addresses issues of implementation including data privacy, model interpretability, and real-time deployment concerns.

Monitoring chicken health is essential for maintaining the production efficiency of poultry farms and meeting the demand for poultry products. Previous studies have explored various methods, including utilizing sound, behaviour, and the shape of the chickens, as well as the conditions of their droppings, to assess chicken health. In this research, we monitor chicken droppings as a reliable indicator of chicken health. We develop an automated system for detecting chicken droppings and identifying health conditions, specifically in open house poultry farms in Malaysia. Open poultry houses are the most common design in Malaysia due to their lower construction and maintenance costs, a more natural environment for the chickens, and greater space to roam. However, the design of open poultry houses, which utilizes evenly gapped wood slat flooring, compounds the problem of automatically distinguishing new droppings from dirty flooring. In our work, data consisting of chicken dropping images from a poultry farm in Malaysia were collected for analysis. We used the YOLOv5n algorithm for detecting chicken droppings and distinguishing between healthy and sick chickens based on observable features such as the colour and shape of their droppings. Our proposed architecture, which used the YOLOv5n algorithm, can accurately detect chicken droppings and classify them into three health classes (coccidiosis, healthy, and other unhealthy), with an accuracy rate of up to 94.9%. By leveraging advanced computer vision techniques, poultry farmers can benefit from timely and accurate health assessments, leading to improved productivity and animal welfare in open house poultry farming systems.

The relationship between individuals and their thermal environment is pivotal not only for comfort but also for health and productivity. Thermal comfort, as defined by ASHRAE, reflects an individual's satisfaction with their ambient thermal conditions and can be gauged using the ASHRAE scale. In the past, traditional thermal comfort prediction models such as the Predicted Mean Vote (PMV) were used to evaluate thermal comfort. Nevertheless, the emergence of machine learning provides a more dynamic approach to predict thermal comfort of occupants. However, the subjective nature of thermal comfort introduces data ambiguities challenge which lead to the existence of outliers. Moreover, data imbalances within the dataset can cause the machine learning models to not learn the minority class effectively, resulting in the deterioration of the model. This research has developed an enhanced thermal comfort prediction model to predict the occupant’s thermal comfort by leveraging the outlier detection technique and synthetic data generator, particularly the Isolation Forest and SMOTE. The experiment showed that the proposed model is able to achieve an accuracy of 74.94%. This exhibited a slight improvement compared to the findings in prior research of using Random Forest prediction model.

This project proposes an operational predictive solution that assists in predicting the flood occurrence at Kota Belud in Sabah, particularly in estimating and predicting the water level, and affected area. It is motivated by the work under the Security And Integrated Flood Operation Network at Kota Belud (S.A.I.F.O.N@Belud). To anticipate the likelihood of a flood, this solution uses machine learning. The suggested method needs input from the S.A.I.F.O.N@Belud system's current sensors. The created solution can assist in determining forecast lead times that enable the authorities to issue prior notice in order to tackle the impending flood catastrophe. The algorithm consists of a Long Short-Term Memory (LSTM) neural network model that can predict the water level of the Tempasuk River in Sabah for the next 30 minutes. The predicted value is then fed into a rule-based flood risk model to predict the flood risk status in the next 30 minutes. The LSTM prediction model developed using data from S.A.I.F.O.N@Belud dataset yields an average of root mean squared error and the mean absolute error of 0.08 and 0.03 respectively. Whereas, the rule-based flood risk model achieves an overall accuracy of 98.18%. The proposed model lays the foundation in expanding the S.A.I.F.O.N@Belud to prediction phase beyond the real time monitoring.

The rapid advancements in deep learning have revolutionized the field of computer vision. However, despite the significant progress in computer vision, there remains a scarcity of research focused on utilizing this technology for distance estimation. Exploring such studies can bring immense convenience to people, especially in applications like anomaly object detection. On that account, this research proposes an improved detection model based on You Only Look Once version 8 (YOLOv8) namely YOLOv8-CAW, which is capable of both detecting target objects and accurately calculating their distances. The proposed method involves incorporating the Coordinate Attention and Wise-IoU into the YOLOv8 network, enhancing the detection accuracy. Combined with the distance estimation algorithm, results in a comprehensive output that includes both detection results and calculated distances. At the end of the experiment, a substantial improvement in performance metrics was observed, the model achieved increases in recall (0.4%), precision (2.2%), and Mean Average Precision (mAP) (1.5%) within the 0.5 to 0.95 threshold range, while maintaining inference speeds similar to the baseline model in PASCAL VOC dataset. Besides that, distance estimation achieved an approximate average accuracy of 90% which shows the results are highly encouraging and promising. The successful integration of computer vision and distance estimation opens new possibilities for practical applications, showcasing the potential of this approach in real-world scenarios.