Floods, common natural disasters in Indonesia, bring substantial damage and economic losses. Jakarta, Indonesia's capital, faces frequent flooding, mainly due to the overflowing Ciliwung River. Effective flood prevention, like timely warnings, is crucial to reduce losses. Yet, the existing flood early warning system managed by the Ciliwung-Cisadane River Center has limitations. Its hydrological models for short-term predictions lack accuracy, leading to prolonged computational times. To enhance this system, a machine learning approach was introduced for better water level prediction. This model utilizes rainwater and runoff data from four river segments upstream. Two neural network methods, Adaptive Neuro-Fuzzy Inference System (ANFIS) and Recurrent Neural Network-Long Short-Term Memory (RNN-LSTM), were compared. RNN-LSTM outperformed ANFIS, displaying lower error rates and faster computation times. It excelled in predictive capabilities for three of four river segments, evident through improved Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE). Despite a minor setback in the third segment's performance, both methods scored well in the coefficient of determination (R2). RNN-LSTM emerged superior due to its minimal error and efficient computational speed, making it the preferred choice for water level prediction. Additionally, RNN-LSTM showcases enhanced predictive ability concerning water level fluctuations, as reflected in its larger standard deviation.
Topic popularity prediction has always been a popular research field. In previous studies, traditional methods are mostly used to predict the popularity of topics characterized by long text, long period, and large amounts of data. These models do not apply to hot topics generated on social network platforms. Hot topics on social platforms are characterized by rapid generation, development, explosion, and vanishing, resulting in topics often disappearing within a few days or even hours, significantly reducing the amount of data available. According to these features, this paper collects blogs related to the hot topic “2023 Hangzhou Asian Games” on Sina Micro-blog, and designs a hotness formula to measure the topic's popularity within a certain period. Subsequently, two baseline models were trained by combining convolutional neural networks (CNN) and long short-term memory (LSTM). The prediction results are compared using two evaluation metrics, RMSE and MAE. The model proposed in this paper outperforms the two baseline models in prediction.
In order to promote viability of the circular economy (CE), circular design in waste management, needs to be highlighted. This study proposes first, situating circular economy design/ development in Value Sensitive Design to increase empathy, and open (instead of closed) innovations; second, incorporating feasibility testing, via SWOT analysis, chain-of-effects mind map and chain-of- effects prioritization, mediated and moderated by economic factors; third, gamifying awareness/contributions, authenticating via blockchain, and analysing/monitoring via a dashboard; fourth, triangulating 3 HCI instruments, to derive data-driven bigger pictures/insights into the technology acceptance of the Figma prototype. By building on open standards and platforms, over time, real-time what-if scenarios and analyses, and knowledge bases can develop; increasing opportunities for business expansion, and better quality of life. Findings from alpha-beta testing are promising, towards enhancing interconnections among (eco)systems, and sustainability/growth. However, since sample size is small, findings cannot be generalized.
This paper examines the comparative effectiveness of three distinct Flipped Classroom (FC) teaching methods in the context of Power and Electrical (PE) engineering education: Traditional Flipped Classroom (TFC), Video-Assisted Teaching Methods (VAFC), and Virtual Reality Flipped Classroom (VRFC). The study incorporates valuable feedback from students who experienced these methods and provides an evaluation of their perceptions. A cross-over methodology for comparing these FC methods is presented. The VAFC method is perceived to be an effective approach and preferred over TFC and VRFC by students. The paper also provides Teachers' perception in implementing above FC methods, which outlines a process for developing VAFC and VRFC teaching resources. Operational difficulties in implementing such methods are highlighted, especially in the context of fast-paced, privately-catered undergraduate education sector striving for implementing non-conventional teaching methods.
In the context of Kun Khmer, a traditional combat technique in Cambodia, the challenge of correcting trainees' postures and movements is significant due to its impact on performance and outcomes. Moreover, incorrect movements during training can lead to physical injuries. This paper presents a machine learning-based system that aims to recognize Kun Khmer postures and potentially facilitate posture correction. The proposed solution utilizes anatomical landmarks and explores multiple classifiers, including Logistic Regression, Random Forest, Gradient Boosting, and Support Vector Machine. Experimental results demonstrate a recognition rate of 96.32% using the Support Vector Machine classifier, with an average recall and precision of 96.22% and 96.32% respectively.
Passive Optical Network (PON) technology is a cost-effective solution for delivering high-speed internet services to residential areas using optical fiber in access networks. The introduction of Next Generation Passive Optical Network 2 (NGPON2) with point-to-point wavelength division multiplexing (WDM) technology has addressed the increasing demand for higher bandwidth. This paper presents a comprehensive analysis of related work in the field, showcasing previous studies on PON coexistence and the challenges they addressed. In addition, a design using Time Division Multiplexing (TDM) and Time Wavelength Division Multiplexing (TWDM) systems is proposed in this paper to demonstrate an effective coexistence of multiple PON technologies. Simulation findings show the system performances that were evaluated at various fiber distances, downstream bit rates. It also identifies areas for future research, such as investigating advanced coexistence techniques, emerging PON technologies, and collaboration with industry partners for validation.
This study aims to inspect a new approach to improve the sidelobe level (SLL) suppression of the radiation pattern in Wireless Sensor Networks (WSNs). WSNs are crucial for various applications that require monitoring and collecting data from multiple sources. One of the challenges in WSNs is the limited transmission range of the sensor nodes, where WSNs often rely on multi-hop transmission. Multi-hop transmission leads to a significant increase in the high transmission power in WSNs. Collaborative beamforming (CB) is proposed to improve communication efficiency through beamforming. In this paper, CB creates a virtual antenna array in a circular arrangement, and the beam is then directed in the desired direction. Then, an Imperialist Competition Algorithm (ICA), is employed to find the optimum solution that suppresses the SLL. ICA is used to optimize inter-element spacing of 10, 12, and 16 circular arrays. The proposed method is compared to the conventional circular array (UCA) and another algorithm. i.e., Backtracking Search Algorithm (BSA). Compared to UCA and Circular BSA, Circular ICA exhibits an overall improvement in SLL suppression of 111.51% and 24.2%, respectively.
This paper presents a comprehensive analysis of climate and environmental risks (CERs) and their significant financial implications. The existing scientific literature lacks adequate recognition and understanding of the financial impacts of CERs, particularly on financial services supporting entities exposed to these risks. The urgency to address these risks becomes paramount for achieving net-zero. To bridge this gap, we propose an AI-enabled framework, Net0Chain, designed to predict and manage the risks associated with climate change. Net0Chain integrates the COBIT 2019 IT Governance model to realize benefits and optimize resources, offering recommendations for achieving net-zero emissions. By implementing a circular economy model, the framework aims to guide financial sectors in translating physical and transition scenarios of climate change into actionable financial strategies. This paper underscores the importance of proactive and informed financial management in the face of growing climate challenges, advocating for a robust approach that aligns with global climate and energy transition targets.
With the surge in artificial intelligence (AI) adoption by Small and Medium-sized Enterprises (SMEs), ensuring their safety, fairness, and operational assurance has become paramount. Since many SMEs operate with limited resources, they face unique challenges in ethically and securely deploying AI systems. This research delves into the core principles of AI governance, risk management, and testing, specifically tailored for SMEs, emphasising making these concepts accessible and understandable. Through collaborative efforts, including interactive workshops, meetings and surveys with twenty SME participants, we identified vital AI application areas and challenges and conceptualised an evaluation tool leveraging explainable AI. This tool assesses AI-driven systems' robustness, potential biases, and other software and hardware vulnerabilities It also addresses ethical considerations and legal compliance, emphasising establishing trust and accountability with stakeholders as a foundation for successful AI integration. In conclusion, the paper presents a pilot study that conducts a risk analysis of prevalent AI applications, specifically AI-driven language models, for SMEs. This study illustrates how the proposed evaluation tool will integrate risk levels across different application domains.
This paper explores the application of Chat Generative Pretrained Transformer (ChatGPT) in the healthcare domain, introducing a sentiment analysis model to evaluate ChatGPT-related tweets in healthcare contexts. The study aims to uncover predominant sentiments, thematic content, and diverse perspectives concerning ChatGPT's integration into healthcare, utilizing an extensive dataset from Twitter comprising 10,330 healthcare-related tweets. Leveraging advanced Natural Language Processing (NLP) techniques, we systematically categorized topics and emotional content within these tweets. Additionally, we conducted a comprehensive analysis of frequently occurring words in tweets expressing positive and negative sentiments. The findings reveal that the majority of healthcare-related ChatGPT tweets express either positive or negative sentiments, with a minor proportion conveying neutral viewpoints. Furthermore, to enhance our comprehension of sentiment dynamics in healthcare discussions involving ChatGPT, we applied four machine learning classifiers Support Vector Machine, K-Nearest Neighbors, Naive Bayes and Random Forest. Remarkably, the SVM classifier demonstrated the highest accuracy at 85.6%, affirming its efficacy in healthcare sentiment analysis. In summary, this research sheds light on prevailing sentiments and perspectives regarding ChatGPT in the healthcare sector, highlighting its predominantly positive and neutral reception on platforms like Twitter. Additionally, the success of SVM as a sentiment analysis tool underscores its potential for discerning sentiments in healthcare-related ChatGPT discussions, contributing to ongoing debates on AI integration in healthcare and guiding future endeavors in this evolving field.
Congenital heart disease (CHD) is a worldwide common congenital disability, and according to a recent WHO report, more than 60,000 babies are born with CHD yearly. Furthermore, total death rates due to CHD reached 16.49% of the total baby deaths. CHD has thirty-five types with various abnormalities in the heart, including abnormal, incomplete, or some parts missing in the heart. Similarly, there are leaky valves and holes in the chamber partitions of the heart. A few types of CHD can be detected during pregnancy and before birth. Patients with CHD need to be diagnosed early, whereas many children die undiagnosed due to CHD because of a lack of awareness about CHD in developing countries. Moreover, in most cases, if CHD is diagnosed in the early stage of life, most people can not afford the cost of CHD surgeries and other procedures for their kids. Recognizing this gap, we propose a deep learning-based model we named the Cardiac Deep Learning Model (CDLM) that can detect this abnormality easily and in the early stage of life with CT-scanned images. We use a segmentation model which segments the four chambers of the heart and then followed by the blood pool stage. We apply a graph-matching method to extract connection information and to know the categories of all the vessels. For experimental work, we use a publically available dataset having 68 CT images of the heart, which covers 14 different kinds of abnormalities. Our proposed method reported better results as compared to the results reported previously.
Elderly standing imbalance is a critical public health concern, demanding robust and accurate detection techniques for improved safety and well-being. In this paper, we propose a novel method employing unsupervised learning and Denoising Autoencoder with Multi-Layer Perceptron networks, along with a custom adaptive Huber loss function and activation function, to classify standing states in elderly individuals. The existing Standing imbalance detection research includes difficulties such as addressing irregularities in pressure sensor data, largely stressing binary classification due to algorithmic efficiency considerations while dealing with heavy-tailed data. The approach utilizes open-source smart insole datasets, capturing left and right foot pressure data. The ensemble model DAE-MLP efficiently captures the temporal dynamics of the imbalance scores produced using the Noise-resilient robust mean estimator, enabling accurate and robust classification. This method adapts to varied degrees of data imbalance, resulting in more accurate learning. Through comprehensive evaluations, our method achieves an overall accuracy of 94 percentage on a test dataset with 53 instances. This approach serves as a proactive standing imbalance detection system for the elderly, enhancing safety and quality of life by identifying and addressing standing imbalance risks. Our research introduces an innovative solution, paving the way for advancements in elderly healthcare and safety, reducing the risk of falls and related injuries.
Widespread usage of the internet and digital communication technologies, multimedia data may be vulnerable to piracy as unauthorized users can easily alter multimedia data such as images. In this study, IWT and SVD based on human visual characteristics are proposed, to embed the watermark in the image, making it more resistant to attacks while low distortion on image quality. The watermarked image is then embedded with authentication bits obtained from the image content of 3x3 block pixels. The authentication bits are embedded into the LSB of pixel to ensure that RGB images can be authenticated if there is any tampering attacks. Various geometrical, compression, image processing attacks and tampering attacks are used to test the effectiveness of the proposed approach. Hence, the test results are compared to other dual watermarking systems. The results of these experiments show that our approached can accurately detect tampering in most cases, even when the changes made to the image affect a small number of blocks. Additionally, we found that our approach provides high robustness, even when image processing attacks are extensive. The results demonstrate the PSNR value of about 48 dB, effectiveness of our authentication approach obtained accuracy of 0.98 for tampering attacks and show that it is a promising technique for robustness of embedded copyright.
Pangasius catfish farming is a significant aquaculture industry in Southeast Asia, expected to make up 80% of global production in 2022, understanding the dynamics of factors affecting fish feeding behaviour is crucial. A 60-day experiment in Cambodia's cages aimed to establish an optimal feeding strategy for Pangasius fish farming, factoring in variables like weather conditions (sunny, cloudy, rainy), pH levels, temperature, and dissolved oxygen. We address this investigation through a critical research question: What should be the optimal feed intake rate of Pangasius fishes under varying conditions? Literature suggests that feed intake is influenced by temperature and sunlight, with higher intake in warm, sunny conditions, but excess sunlight can induce stress. Cloudy and rainy weather tends to reduce feeding and growth. Our data analysis revealed that factors like temperature, pH, and dissolved oxygen affect optimal feeding rates. On sunny days, higher feeding rates are feasible, but caution is needed to avoid overfeeding. Conversely, on cloudy or rainy days, it's advisable to reduce feeding rates to prevent waste accumulation and maintain water quality. This factors-based feeding strategy can enhance feed efficiency, profitability, and sustainability in Pangasius aquaculture.
A switched reluctance machine (SRM) has some advantages. Due to the absence of permanent magnets on its rotor, operating it as a generator requires two stages namely excitation stage and generation stage. Under the excitation stage, batteries are needed to excite the stator winding, this event must occur in proper commutation angles. These angles will determine the energy which will be produced during generation stage. SRM stator inductance will vary depend on the relative position between rotor and stator. A few moments before maximum value of the stator inductance, excitation to the stator winding must be given while turning off the stator excitation must be done during the negative slope of the stator inductance. In this paper, a method to generate precise commutation angles for SRM is proposed. This method is based on the input capture facility provided by dsPIC30F4012 as the core of the digital control. To verify the analysis, simulation and experimental works were done.
Facial Beauty Prediction (FBP) is a computer vision task of quantifying the beauty of a face. Several solutions to this problem have benefitted immensely from the recent developments in deep learning. However, the majority of current methods train machine learning models to purely predict mean beauty scores, treating FBP solely as a regression task. In addition, deep learning based FBP approaches so far use transfer learning from models trained on general classification tasks such as ImageNet. We propose fine-tuning an ensemble of convolutional neural network (CNN) models originally trained on face verification tasks using a variety of loss functions such as Earth Mover's Distance (EMD) based loss. With this approach, our method can predict the entire beauty score distribution rather than just the mean, and the predicted mean scores have a higher Pearson Correlation (PC) compared to the ground truth scores. This method achieves state of the art results on the MEBeauty dataset in terms of mean absolute error, root mean squared error and PC between the predicted and the ground truth mean scores.
This research aims to study the relationship between financial inclusion efficiency (FIE) and poverty score. The data for this research includes 28states and 3 union territories of India. A data envelopment analysis (DEA) is a popular and effective tool revealed by literature research studies to assess countries' financial inclusion levels. This study used a panel data model with DEA wherethe Poverty Score (PS) is presented as the dependent variable, and the independent variables selected are Financial Inclusion Efficiency (FI_CRS), Number of ATMs (No_ATM), and Literacy Rate (Lit_rt). From the optimal weights of the DEA model, descriptive statistics, correlation matrix scores, and endogeneity test results, the results of the study show that there is a significant impact on poverty scores by financial inclusion efficiency. The Financial Inclusion Efficiency (FI_CRS) is significant and negative which signifies that there is a negative relationship between Financial Inclusion Efficiency (FI_CRS) and Poverty Score (PS). Thus, this research shows that the rise in fie resulted in a decrease in the Poverty Score of 28 Indian states and 03 union territories.
Solar power generation is becoming increasingly popular to meet the continuously growing demand for electricity, as it falls under the category of renewable energy sources that are abundant and inexhaustible. This type of power generation converts direct sunlight into electrical energy using the photovoltaic effect. However, solar panels experience a decrease in output efficiency due to various factors, one of which is shading. In solar power systems, shading is the most significant factor that affects power output and has a considerable potential to disrupt photovoltaic performance. Shading can occur due to weather conditions such as the movement of the sun or clouds, as well as environmental factors like trees or buildings. This research aims to examine the impact of partial shading on the output power of solar panels using an on-grid inverter. Several experiments were conducted by applying shading effects in various patterns to the photovoltaic array to determine their effects on power output. Data collection was performed using two data acquisition devices to minimize errors in the collected data.
Companies need to be aware how Industry 4.0 technologies fit to their working and effect their business processes, what modifications are required before implementing them. Digital maturity and readiness models have been created to meet these requirements. This paper shows how one part of the model developed in the IPAMS project can be mapped into ontology and how ontology development can contribute to automatize the evaluation of company digital readiness.
This conference paper presents the design and development of a large-scale hexapod robot inspired by the agile locomotion of rabbits, utilizing the Steed-Beast Mechanism. Inspired by the remarkable biomechanics of rabbit legs, known for their swift and efficient hopping motion, this research seeks to replicate the dynamic gait of these animals to create a versatile robotic platform capable of navigating diverse terrains. The design process encompasses the biomimetic structure of rabbit legs, precise articulation of joints, powerful actuators, and adaptive feet for enhanced grip. The Steed-Beast Mechanism further enhances the robot's mobility and stability. Through rigorous testing and iterative refinement, we have successfully demonstrated the robot's ability to replicate the unique hopping gait of rabbits. This conceptual design holds great potential for applications in agriculture, and exploration, and its adaptability marks a significant advancement in large-scale hexapod robotics. This paper not only presents the design and implementation of the robot but also discusses its significance in the broader context of biomechanically inspired robotics.
