A Bi-level Neuro-Fuzzy System Soft Computing for Reservoir Operation

Reservoir operation studies purely based on the storage level, inflow, and release decisions during dry periods only fail to serve the optimal reservoir operation policy design because of the fact that the release decision during this period is highly dependent on wet season water conservation and flood risk management operations. Imperatively, the operation logic in the two seasons are quite different. If the two operations are not sufficiently coordinated, they may produce poor responses to the system dynamics. There are high levels of uncertainties on the model parameters, values and how they are logically operated by human or automated systems. Soft computing methods represent the system as an artificial neural network (ANN) in which the input- output relations take the form of fuzzy numbers, fuzzy arithmetic and fuzzy logic (FL). Neuro-Fuzzy System (NFS) soft computing combine the approaches of FL and ANN for single purpose reservoir operation. Thus, this study proposes a Bi-Level Neuro-Fuzzy System (BL-NFS) soft computing methodology for short and long term operation policies for a newly inaugurated irrigation project in Gidabo Watershed of Main Ethiopian Rift Valley Basin. Keywords: Bankruptcy rule, BL-NFS, Reservoir operation, Sensitivity analysis, Soft computing, Water conservation.

An Intelligent Vice Cluster Head Election Protocol in WSN

Wireless sensor networks (WSNs) has a practical ability to link a set of sensors to build a wireless network that can be accessed remotely; this technology has become increasingly popular in recent years. Wi-Fi-enabled sensor networks (WSNs) are used to gather information from the environment in which the network operates. Many obstacles prevent wireless sensor networks from being used in a wide range of fields. This includes maintaining network stability and extending network life. In a wireless network, sensors are the most essential component. Sensors are powered by a battery that has a finite amount of power. The battery is prone to power loss, and the sensor is therefore rendered inoperative as a result. In addition, the growing number of sensor nodes off-site affects the network's stability. The transmission and reception of information between the sensors and the base consumes the most energy in the sensor. An Intelligent Vice Cluster Head Selection Protocol is proposed in this study (IVC LEACH). In order to achieve the best performance with the least amount of energy consumption, the proposed hierarchical protocol relies on a fuzzy logic algorithm using four parameters to calculate the value of each node in the network and divides them into three hierarchical levels based on their value. This improves network efficiency and reliability while extending network life by 50 percent more than the original Low Energy Adaptive Clustering Hierarchy protocol. Keywords: Wireless Sensor Networks, Sensors, Communication Protocol, Fuzzy logic, Leach protocol.

Arrhythmia Classification Using One Dimensional Conventional Neural Network

Arrhythmia is a heart condition that occurs due to abnormalities in the heartbeat, which means that the heart's electrical signals do not work properly, resulting in an irregular heartbeat or rhythm and thus defeating the pumping of blood. Some cases of arrhythmia are not considered serious, while others are very dangerous, life-threatening, and cause death in a short period of time. In the clinical routine, cardiac arrhythmia detection is performed by electrocardiogram (ECG) signals. The ECG is a significant diagnosis tool that is used to record the electrical activity of the heart, and its signals can reveal abnormal heart activity. However, because of their small amplitude and duration, visual interpretation of ECG signals is difficult. As a result, we present a significant approach for identifying arrhythmias using ECG signals. In this study, we proposed an approach based on Deep Learning (DL) technology that is a framework of nine-layer one-dimension Conventional Neural Network (1D CNN) for classifying automatically ECG signals into four cardiac conditions named: normal (N), Atrial Premature Beat (APB), Left Bundle Branch Block (LBBB), and Right Bundle Branch Block (RBBB). The practical test of this work was executed with the benchmark MIT-BIH database. We achieved an average accuracy of 99%, precision of 98%, recall of 96.5%, specificity of 99.08%, and an F1-score of 95.75%. The obtained results were compared with some relevant models, and they showed that the proposed framework outperformed those models in some measures. The new approach’s performance indicates its success. Also, it has been shown that deep convolutional neural networks can be used efficiently in automated detection and, therefore, cardiovascular disease protection as well as help cardiologists in medical practice by saving time and effort. Keywords: 1-D CNN, Arrhythmia, Cardiovascular Disease, Classification, Deep learning, Electrocardiogram(ECG), MIT-BIH arrhythmia database.

Artificial Intelligence Approach in Multiclass Diabetic Retinopathy Detection Using Convolutional Neural Network and Attention Mechanism

The number of people around the world who have diabetes is about 422 million. Diabetes seriously affects the blood vessels in the retina, a disease called diabetic retinopathy (DR). The ophthalmologist examines signs through fundus images, such microaneurysm, exudates and neovascularisation and determines the suitable treatment for patient based on the condition. Currently, doctors require a long time and professional skills to detect DR. This study aimed to implement artificial intelligence (AI) to resolve the lack of current methods. This study implemented AI for detecting and classifying DR. AI uses deep learning, such the attention mechanism algorithm and AlexNet architecture. The attention mechanism algorithm focuses on detecting the pathological area in the fundus images, and AlexNet is used to classify DR into five levels based on the pathological area. This study also compared AlexNet architecture with and without attention mechanism. We obtained 344 fundus images from the Kaggle dataset, which contains normal, mild, moderate, severe and proliferative DR. The highest accuracy in this study is up to 91% and used the attention mechanism algorithm and AlexNet architecture. The experiment shows that our proposed method can provide results that can detect the pathological areas and effectively classify DR. Keywords: Artificial intelligence, Diabetic Retinopathy, Attention Mechanism, AlexNet

Bilingual Text Classification in English and Indonesian via Transfer Learning using XLM-RoBERTa

With the large amount of text information circulating on the internet, there is a need of a solution that can help processing data in the form of text for various purposes. In Indonesia, text information circulating on the internet generally uses 2 languages, English and Indonesian. This research focuses in building a model that is able to classify text in more than one language, or also commonly known as multilingual text classification. The multilingual text classification will use the XLM-RoBERTa model in its implementation. This study applied the transfer learning concept used by XLM-RoBERTa to build a classification model for texts in Indonesian using only the English News Dataset as a training dataset with Matthew Correlation Coefficient value of 42.2%. The results of this study also have the highest accuracy value when tested on a large English News Dataset (37,886) with Matthew Correlation Coefficient value of 90.8%, accuracy of 93.3%, precision of 93.4%, recall of 93.3%, and F1 of 93.3% and the accuracy value when tested on a large Indonesian News Dataset (70,304) with Matthew Correlation Coefficient value of 86.4%, accuracy, precision, recall, and F1 values of 90.2% using the large size Mixed News Dataset (108,190) in the model training process. Keywords: Multilingual Text Classification, Natural Language Processing, News Dataset, Transfer Learning, XLM-RoBERTa

Potential Security Vulnerabilities of the IEEE 802.15.4 Standard and a Proposed Solution Against the Dissociation Process

Many different networks that rely on short-distance wireless technology for their functions utilize the IEEE 802.15.4 Standard, especially in the case of systems that experience a low level of traffic. The networks using this standard are typically based on the Low-Rate Wireless Personal Area Network, herein called the LR-WPAN; this network is used for the provision of both the physical layer, herein referred to as the PHY, and the media access control, herein abbreviated as the MAC. There are four security features in the IEEE 802.15.4 Standard that are designed to ensure the safe and secure transmission of data through the network. Disconnection from the network is managed and controlled by the message authentication code, herein referred to as the MAC, while the coordinator personal area network, herein abbreviated as the PAN, is also able to trigger the disconnection. However, the process of disconnection from the network is one area of vulnerability to denial-of-service attacks, herein referred to as DoS; this highlights a major shortcoming of the IEEE 802.15.4 Standard’s security features. This paper is intended to contribute to the improvement of security for the IEEE network by conducting a specific and in-depth review of available literature as well as conducting an analysis of the disassociation process. In doing so, potential new threats will be highlighted, and this data can be used to improve the security of the IEEE 802.15.4 Standard. Overall, in this paper, the role of the Castalia tool in the OMNET++ environment is analysed and interpreted for these potential new threats. Also, this paper proposes a solution to such threats to improve the security IEEE 802.15.4 disassociation process. Keywords: Disassociation vulnerability of IEEE 802.15.4 Standard, DoS attack, IoT security.

Web Scraping or Web Crawling: State of Art, Techniques, Approaches and Application

Web scraping or web crawling refers to the procedure of automatic extraction of data from websites using software. It is a process that is particularly important in fields such as Business Intelligence in the modern age. Web scrapping is a technology that allow us to extract structured data from text such as HTML. Web scrapping is extremely useful in situations where data isn’t provided in machine readable format such as JSON or XML. The use of web scrapping to gather data allows us to gather prices in near real time from retail store sites and provide further details, web scrapping can also be used to gather intelligence of illicit businesses such as drug marketplaces in the darknet to provide law enforcement and researchers valuable data such as drug prices and varieties that would be unavailable with conventional methods. It has been found that using a web scraping program would yield data that is far more thorough, accurate, and consistent than manual entry. Based on the result it has been concluded that Web scraping is a highly useful tool in the information age, and an essential one in the modern fields. Multiple technologies are required to implement web scrapping properly such as spidering and pattern matching which are discussed. This paper is looking into what web scraping is, how it works, web scraping stages, technologies, how it relates to Business Intelligence, artificial intelligence, data science, big data, cyber securityو how it can be done with the Python language, some of the main benefits of web scraping, and what the future of web scraping may look like, and a special degree of emphasis is placed on highlighting the ethical and legal issues. Keywords: Web Scraping, Web Crawling, Python Language, Business Intelligence, Data Science, Artificial Intelligence, Big Data, Cloud Computing, Cybersecurity, legal, ethical.

Embedding from Language Models (ELMos)- based Dependency Parser for Indonesian Language

The goal of dependency parsing is to seek a functional relationship among words. For instance, it tells the subject-object relation in a sentence. Parsing the Indonesian language requires information about the morphology of a word. Indonesian grammar relies heavily on affixation to combine root words with affixes to form another word. Thus, morphology information should be incorporated. Fortunately, it can be encoded implicitly by word representation. Embeddings from Language Models (ELMo) is a word representation which be able to capture morphology information. Unlike most widely used word representations such as word2vec or Global Vectors (GloVe), ELMo utilizes a Convolutional Neural Network (CNN) over characters. With it, the affixation process could ideally encoded in a word representation. We did an analysis using nearest neighbor words and T-distributed Stochastic Neighbor Embedding (t-SNE) word visualization to compare word2vec and ELMo. Our result showed that ELMo representation is richer in encoding the morphology information than it's counterpart. We trained our parser using word2vec and ELMo. To no surprise, the parser which uses ELMo gets a higher accuracy than word2vec. We obtain Unlabeled Attachment Score (UAS) at 83.08 for ELMo and 81.35 for word2vec. Hence, we confirmed that morphology information is necessary, especially in a morphologically rich language like Indonesian. Keywords: ELMo, Dependency Parser, Natural Language Processing, word2vec

Adaptive Multi-Applications Cryptographic System

Cryptography is the core method utilized to protect the communications between different applications, terminals, and agents distributed worldwide and connected via the internet. Yet, with the distribution of the low-energy and low-storage devices, in the Internet-of-Things (IoT), the cryptography protocols cannot be implemented because of the power constraints or because the implementation is beyond the time constraints that hindered their usability of these protocols in real-time critical applications. To solve this problem, an Adaptive Multi-Application Cryptography System is proposed in this paper. The proposed system consists of the requirements identifier and the implementer, implemented on the application and transportation layer. The requirement identifier examines the header of the data, determines the underlying application and its type. The requirements are then identified and encoded according to four options: high, moderate, low, and no security requirements. The inputs are processed, and ciphertext is produced based on the identified requirements and the suitable cryptography algorithm. The results showed that the proposed system reduces the delay by 97% relative to the utilized algorithms' upper-bound delay. Keywords: Cryptography, symmetric key encryption, block cipher, delay and performance, quantum computing.

Employing of Object Tracking System in Public Surveillance Cameras to Enforce Quarantine and Social Distancing Using Parallel Machine Learning Techniques

Like many countries, Jordan has resorted to lockdown in an attempt to contain the outbreak of Coronavirus (Covid-19). A set of precautions such as quarantines, isolations, and social distancing were taken in order to tackle its rapid spread of Covid-19. However, the authorities were facing a serious issue with enforcing quarantine instructions and social distancing among its people. In this paper, a social distancing mentoring system has been designed to alert the authorities if any of the citizens violated the quarantine instructions and to detect the crowds and measure their social distancing using an object tracking technique that works in real-time base. This system utilises the widespread surveillance cameras that already exist in public places and outside many residential buildings. To ensure the effectiveness of this approach, the system uses cameras deployed on the campus of Al-Zaytoonah University of Jordan. The results showed the efficiency of this system in tracking people and determining the distances between them in accordance with public safety instructions. This work is the first approach to handle the classification challenges for moving objects using a shared-memory model of multicore techniques. Keywords: Covid-19, Parallel computing, Risk management, Social distancing, Tracking system.