Bayesian Classifier Research Articles

State Evaluation of Electrical Equipment in Substations Based on Data Mining

This paper explores the combination of a data mining-based state evaluation method for electrical equipment in substations, analyzing the effectiveness and accuracy. First, a Gaussian mixture model is applied to fit all raw data of electrical equipment. The Expectation Maximization algorithm summarizes the data distribution characteristics and identifies outliers. The a priori algorithm is then employed for data mining to derive frequent itemsets and association rules between equipment quality and measurement data. For new equipment samples, conditional probabilities of each feature are independently calculated and combined to classify and evaluate equipment quality. The results suggest that equipment reliability in smart substations can be inferred from historical and real-time operational data using improved association rule algorithms and Naive Bayes classifiers. Finally, the proposed method was applied to analyze statistical data from a 110 kV substation of a power supply company. The states prediction accuracy exceeded 95% when compared with actual equipment quality. The effectiveness evaluation metrics demonstrated that this method outperforms single-category algorithms in terms of accuracy and discrimination ability.

A PREDICTION MODEL FOR STROKE BASED ON MACHINE LEARNING ALGORITHMS

In this modern era, people are working hard to meet their physical needs and non-effective their ability to spend time for themselves which leads to physical stress and mental disorder. Many reports state that stroke is caused when blood flow to a part of the brain is stopped abruptly. Without the blood supply, the brain cells gradually die, and disability occurs depending on the area of the brain affected. According to the World Health Organization (WHO), stroke is the greatest cause of death and disability globally. Early recognition of the various warning signs of a stroke can help reduce the severity of the stroke. In this research work, with the aid of machine learning (ML), several algorithms has been used and evaluated for the long term risk prediction of stroke occurrence. We have collected datasets to analyze data and mining using 8 algorithms of machine learning to predict whether the patient suffers from stroke or not. This paper used a dataset retrieved from kaggle repository, which consists of 12 attributes (Features). This work is implemented using K-Nearest Neighbors (KNN), Naïve Bayes (NB), Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), Decision Tree (TD), Neural Network (NN) and eXtreme Gradient Boosting (XGB) algorithms. Results showed eXtreme Gradient Boosting (XGB) gave the best result with an accuracy of up to 95.14%.

Automatic similarity feature selection for ontology matching with semantic sampling

Ontology provides a formal representation of knowledge as a set of concepts within a domain and the relationships between them. Ontology Matching (OM) identifies similar entities across different ontologies. Similarity Features (SFs) characterize the similarity between two entities from different perspectives, which serves as the foundation of OM method. However, noisy and redundant SFs can obstruct the relevance of useful ones, reducing the quality of matching results. To offer a practical and efficient SF selection, this work proposes a new automatic SF selection framework for OM, which consists of two new components. First, a semantic sampling method is proposed to automatically construct a balanced and representative training dataset, without the need for expert intervention. This method first generates a basic sample set that considers both the representativeness of the concepts and the heterogeneous characteristics of the two ontologies, and then over-samples the minority samples based on their semantic context. Second, a wrapper-based SF selection method is designed, which includes an Adaptive Compact Genetic Algorithm (ACGA) and a Relaxed Naive Bayes (RNB) classifier. The ACGA is able to efficiently identify the SF subset, reducing the computational cost by adaptively maintaining a population probability model based on the diversity and confidence of the SFs. Moreover, the RNB can effectively evaluate the quality of selected SFs by modifying the feature independence assumption with recombination and weighting strategies. Experimental results on the ontology alignment evaluation initiative’s Benchmark demonstrate that our method can efficiently find accurate OM results, and significantly outperform the state-of-the-art matching techniques.

Detection and Analysis of Malicious Software Using Machine Learning Models

The continuous evolution of malware poses a significant challenge in cybersecurity, adapting to technological advancements despite implemented security measures. This paper introduces an innovative approach to enhance the detection of obfuscated malware through the integration of machine learning (ML). Utilizing a real-world dataset of prevalent malware types such as spyware, ransomware, and trojan horses, our study addresses the evolving challenges of cybersecurity. In this study, we evaluate the performance of ML algorithms for obfuscated malware detection using the CIC-MalMem-2022 dataset. Our analysis encompasses binary and multi-class classification tasks under various experimental conditions, including percentage splits and 10-fold cross-validation. The evaluated algorithms include Random Tree (RT), Random Forest (RF), J-48 (C4.5), Naive Bayes (NB), and XGBoost. Experimental results demonstrate the effectiveness of RF, J-48, and XGBoost in achieving high accuracy rates across different classification tasks. NB also shows competitive performance but faces challenges in handling imbalanced datasets and multi-class classification. Our findings highlight the importance of employing advanced ML techniques for enhancing obfuscated malware detection capabilities and provide valuable insights for cybersecurity practitioners and researchers. Future research directions include fine-tuning model hyperparameters, exploring ensemble learning approaches, and expanding evaluation to diverse datasets and real-world scenarios.

A Predictive Model Using Long Short-Time Memory (LSTM) Technique for Power System Voltage Stability

The stability of the operation of the power system is essential to ensure a continuous supply of electricity to meet the load of the system. In the operational process, voltage stability (VS) should be recognized and predicted as a basic requirement. In electrical systems, deep learning and machine learning algorithms have found widespread applications. These algorithms can learn from previous data to detect and predict future scenarios of potential instability. This study introduces long short-term memory (LSTM) technology to predict the stability of the nominal voltage of the power system. Based on the results, the recommended LSTM technology achieved the highest accuracy target of 99.5%. In addition, the LSTM model outperforms other machine learning (ML) and deep learning techniques, i.e., support vector machines (SVMs), Naive Bayes (NB), and convolutional neural networks (CNNs), when comparing the accuracy of the VS forecast. The results show that the LSTM method is useful to predict the voltage of an electrical system. The IEEE 33-bus system indicates that the recommended approach can rapidly and precisely verify the system stability category. Furthermore, the proposed method outperforms conventional assessment methods that rely on shallow learning.

Overall Study in Image Classification Techniques

The era in which we live has become the era of big data, particularly with the spread of the concept of Internet-of-Things (IoT) and artificial-intelligence (AI). Images constitute a large portion of this data that is created and handled. Image classification is the process of identifying and labeling groups of pixels or vectors inside an image using certain criteria. Major image classification techniques divided into two categories which are supervised and unsupervised techniques. This study looks at various research, techniques, and issues of image classification. The focus is on summarizing the main advanced classification strategies and methods such as (K-Mean clustering, Fuzzy measure, Artificial Neural Networks (ANN), Decision-Tree (DT), Support-Vector-Machines (SVMs), Naive Bayes (NB), K-Nearest-Neighbor (KNN), Random-Forest (RF), etc.) that can be utilized after some update to enhance classification accuracy. Also, this study shows the common challenges and solutions in this area.

Exploration of novel human neutrophil elastase inhibitors from natural compounds: Virtual screening, in vitro, molecular dynamics simulation and in vivo study

BackgroundHuman neutrophil elastase (HNE) is an important contributor to lung diseases such as acute lung injury (ALI) or acute respiratory distress syndrome. Therefore, this study aimed to identify natural HNE inhibitors with anti-inflammatory activity through machine learning algorithms, in vitro assays, molecular dynamic simulation, and an in vivo ALI assay. MethodsBased on the optimized Discovery Studio two-dimensional molecular descriptors, combined with different molecular fingerprints, six machine learning models were established using the Naïve Bayesian (NB) method to identify HNE inhibitors. Subsequently, the optimal model was utilized to screen 6925 drug-like compounds obtained from the Traditional Chinese Medicine Systems Pharmacy Database and Analysis Platform (TCMSP), followed by ADMET analysis. Finally, 10 compounds with reported anti-inflammatory activity were selected to determine their inhibitory activities against HNE in vitro, and the compounds with the best activity were selected for a 100 ns molecular dynamics simulation and its anti-inflammatory effect was evaluated using Poly (I:C)-induced ALI model. ResultsThe evaluation of the in vitro HNE inhibition efficiency of the 10 selected compounds showed that the flavonoid tricetin had the strongest inhibitory effect on HNE. The molecular dynamics simulation indicated that the binding of tricetin to HNE was relatively stable throughout the simulation. Importantly, in vivo experiments indicated that tricetin treatment substantially improved the Poly (I:C)-induced ALI. ConclusionThe proposed NB model was proved valuable for exploring novel HNE inhibitors, and natural tricetin was screened out as a novel HNE inhibitor, which was confirmed by in vitro and in vivo assays for its inhibitory activities.

Integrating electro-mechanical impedance data with machine learning for damage detection and classification of blended concrete systems

In recent years, the assessment of structural integrity and durability of concrete systems has significantly advanced with the integration of novel sensing technologies and data-driven techniques. This study introduces an innovative approach that utilizes electro-mechanical impedance (EMI) data acquired through embedded piezo sensors (EPS) to monitor and analyze the condition of blended concrete systems. Moreover, the study involves blended concrete specimens are subjected to controlled damage scenarios under mechanical loadings to establish EPS as an early warning sensor for detecting damage in these systems. Furthermore, the extent of damage under these loadings is comprehensively evaluated using statistical indices and an equivalent stiffness parameter. Additionally, the study focuses on the application of machine learning (ML) algorithms to interpret EMI data for efficient damage classification. ML models, such as decision trees (DT), random forests (RF), k-nearest neighbors (KNN), support vector machines (SVM), and naive Bayes (NB) are trained to identify patterns correlating with different types and extents of structural damage. Among them, the RF classifier exhibits the highest accuracy of 0.91, followed by DT, KNN, SVM, and NB with accuracies of 0.88, 0.86, 0.68, and 0.17 respectively. The integration of EMI data with ML approaches demonstrates significant potential to revolutionize predictive maintenance strategies, enabling precise classification of structural damage and providing critical insights for engineers and stakeholders to proactively address issues, thereby significantly enhancing the safety, durability, and lifespan of concrete infrastructure. This study not only extends the current understanding of EMI data applications but also opens new paths for automated and real-time structural health monitoring.

Sentiment Analysis of PeduliLindungi User Using Naïve Bayes Classifier Algorithm and Support Vector Machine

The Indonesian government is attempting to track the spread of the virus by creating an application named “PeduliLindungi” to deal with the coronavirus's exponential increase in cases across the country. Because it has a feature to disclose the user's location immediately, it is envisaged that this program can reduce the transmission of viruses in monitoring. Indonesians have used the PeduliLindungi, and there are user reviews of both positive and negative experiences. Therefore, to enhance these services, an assessment is required. The text mining method can extract information from users' reviews to collect this data. This method's application additionally uses the Naive Bayes Classifier and Support Vector Machine algorithms, which analyze word associations and do a classification evaluation of the data's accuracy. Based on the two methods' calculations, the NBC algorithm's average classification accuracy was 83.81%, and the SVM algorithm was 93.84%. Following that, discoveries on words that frequently exist or are used by people are obtained through word associations in the sentiment analysis of positive or negative reviews.

A Novel Method of Using Machine Learning Techniques to Protect Clouds Against Distributed Denial of Service (DDoS) Attacks.

The term "cloud computing" describes a method of providing hardware- and software-based services over the internet. This allows users to access their data and apps from any device. The benefits of cloud computing include scalability, virtualization, access to user assets, lower infrastructure costs, and flexibility. However, one drawback is that it is susceptible to distributed denial of service attacks, which occur when multiple computer systems collaborate to target a particular resource, website, or server. Distributed denial of service (DDoS) attacks present a serious risk to computer networks and constitute a major cyber security challenge. This results in a denial of service for end users, as a result of false connection requests, a flood of messages, and twisted packets causing the system to slow down or even crash. Real people and services cannot access cloud computing. The issue of machine learning algorithms for identifying distributed denial of service (DDoS) attacks is explored in this article. In order to identify and defend cloud systems from harmful assaults, we developed a new machine learning approach in this work that is based on transfer learning. This study offers NSL-KDD datasets and two methodologies. There are two types of filters available: the Learning Vector Quantization (LVQ) Filters and the Principal Component Analysis (PCA) method, which reduces dimensionality. The features selected from each method were pooled using Decision Tree (DT), Naïve Bayes (NB), and Support Vector Machine (SVM) to detect distributed denial of service attacks (DDoS). We contrasted the results of several classifications. In terms of attack detection, LVQ-based DT performed better results as compared to other methods.

Identification of medication-related fall risk in adults and older adults admitted to hospital: A machine learning approach

The study aimed to develop and validate, through machine learning, a fall risk prediction model related to prescribed medications specific to adults and older adults admitted to hospital. A case-control study was carried out in a tertiary hospital, involving 9,037 adults and older adults admitted to hospital in 2016. The variables were analyzed using the algorithms: logistic regression, naive bayes, random forest and gradient boosting. The best model presented an area under the curve = 0.628 in the older adult subgroup, compared to an area under the curve (AUC) = 0.776 in the adult subgroup. A specific model was developed for this sample. The gradient boosting model presented the best performance in the sample of older adults (AUC = 0.71). Models developed to predict the risk of falls based on medications specifically aimed at older adults presented better performance in relation to models developed in the total population studied.

Comparison of the impact of dimensionality reduction and data splitting on classification performance in credit risk assessment

Credit risk assessment (CRA) plays an important role in credit decision-making process of financial institutions. Today, developing big data analysis and machine learning methods have marked a new era in credit risk estimation. In recent years, using machine learning methods in credit risk estimation has emerged as an alternative method for financial institutions. The past demographic and financial data of the person whose CRA will be performed is important for creating an automatic artificial intelligence credit score prediction model based on machine learning. It is also necessary to use features correctly to create accurate machine learning models. This article aims to investigate the effects of dimensionality reduction and data splitting steps on the performance of classification algorithms widely used in the literature. In our study, dimensionality reduction was performed using Principal Component Analysis (PCA). Random Forest (RF), Logistic Regression (LR), Decision Tree (DT), Naive Bayes (NB) algorithms were chosen for classification. Percentage splitting (PER, 66–34%) and k-fold (k = 10) cross-validation techniques were used when dividing the data set into training and test data. The results obtained were evaluated with accuracy, recall, F1 score, precision, and AUC metrics. German data set was used in this study. The effect of data splitting and dimension reduction on the classification of CRA systems was examined. The highest ACC in PER and CV data splitting was obtained with the RF algorithm. Using data splitting methods and PCA, the highest accuracy was observed with RF and the highest AUC with NB, with 13 PCs in which 80% of the variance was obtained. As a result, the data set consisting of a total of 20 features, expressed by 13 PCs, achieved similar or higher success than the results obtained from the original data set.

Sentiment Analysis of Joglo Wifi UMKM Service with Naive Bayes Method

In today's digital era, many businesses face challenges in managing and responding to customer feedback and suggestions contained in comments. Joglo Wifi Corner UMKM Business Center & Coffee experiences similar difficulties, where the lack of customer feedback management negatively impacts service quality and customer satisfaction. In this study, we aim to develop and implement a sentiment analysis system that is able to classify customer sentiment into positive, negative, or neutral. The process begins with data collection. customer comments from the official Joglo Wifi Corner website, which are then processed through a preprocessing stage including tokenization, stop word removal, and lemmatization. Then, a Logistic Regression Naive Bayes model is drilled using the processed data to classify sentiment. The evaluation results show that this system achieves an accuracy of 91.67%, with an average precision of 0.646, an average recall of 0.805, and an average F1-Score of 0.701. The implementation of this system provides valuable insights for Joglo Wifi Corner managers in making strategic decisions to improve service quality. By responding more effectively to customer feedback, Joglo Wifi Corner promises that this research can improve customer satisfaction and expand the company's market coverage and identify areas for improvement based on customer feedback, thereby strengthening customer relationships. This approach also uses Agile Development system development methodology to ensure iterative and responsive development to changing user needs.

REAL TIME MONITORING OF AIR POLLUTION USING ARTIFICIAL INTELLIGENCE BASED IOT AND NOVEL SARIMA TECHNIQUE IN SIVAKASI - INDIA

<p style="text-align:justify; margin-bottom:13px"><span style="font-size:11pt"><span style="line-height:200%"><span style="font-family:Calibri,sans-serif"><span style="font-size:12.0pt"><span style="line-height:200%"><span style="font-family:"Times New Roman",serif">Air pollution, a harmful or excessive quantity of pollutants from natural sources and human activities, poses risks to human health, the environment, and ecosystems. AI breakthroughs have allowed for the incorporation of technologies into performance indices, resulting in the development of an AI-based air quality system that evaluates water quality in real time using WHO-defined parameters. This article describes the implementation and planning of AI-based IoT for air pollution tracking and forecasting utilizing AI methodologies, as well as a dashboard on the internet for real-time tracking of air pollutants via Google Cloud servers. Air pollutants such as NO<sub>2</sub>, NO<sub>x</sub>, NH<sub>3</sub>, CO, SO<sub>2</sub>, and O<sub>3</sub> are gathered from IoT sensor nodes in Sivakasi, Tamil Nadu, India, utilizing artificial intelligence algorithms. Individual pollutants are forecasted using time series modeling approaches such as Artificial Neural Network (ANN), Naive Bayes Model, k-nearest neighbour (k-NN), Support Vector Machine (SVM), and Seasonal Autoregressive Interated Moving Average (SARIMA). The data from the IoT sensor node is utilized to train the model, resulting in optimal parameters. The derived model parameters are validated using new, previously unknown data for time. The performances of several Time Series models are examined using performance metrics such as Mean Absolute Error (MAE), coefficient of determination (R<sup>2</sup>), and Root Mean Square Error (RMSE). An AI-based algorithm has been flashed in the Raspberry Pi 3. The present air pollution data and anticipated data are monitored throughout a 7days from 10 p.m. to 4 a.m. using a digital dashboard built in an open-source using Google cloud services. Finally comparing to all above AI based algorithms, SARIMA performed well and h+ad a 95% accuracy level. </span></span></span></span></span></span></p>

An ensemble machine learning-based approach to predict cervical cancer using hybrid feature selection

Cervical cancer has recently emerged as the leading cause of premature death among women. Around 85% of cervical cancer cases occur in underdeveloped countries. There are several risk factors associated with cervical cancer. This study describes a novel predictive model that uses early screening and risk trends from individual health records to forecast cervical cancer patients' prognoses. This study uses machine learning classification techniques to investigate the risk factors for cervical cancer. Additionally, use the voting method to evaluate all models and select the most appropriate model. The dataset used in this study contains missing values and shows a significant imbalance. Thus, the Random Oversampling technique was used as a sampling method. We used Principal Component Analysis (PCA) and XGBoost feature selection techniques to determine the most important features. To predict the accuracy, we used several machine learning classifiers, including Support Vector Machines (SVM), Random Forest (RF), k-nearest Neighbors (KNN), Decision Trees (DT), Naive Bayes (NB), Logistic Regression (LR), AdaBoost (AdB), Gradient Boosting (GB), Multilayer Perceptron (MLP), and Nearest Centroid Classifier (NCC). To demonstrate the efficacy of the suggested model, a comparison of its accuracy, sensitivity, and specificity was performed. We used the Random Oversampling approach along with the Ensemble ML method, hard voting on RF and MLP, and achieved 99.19% accuracy. It is demonstrated that the ensemble ML classifier (hard voting) performs better at handling classification problems when features are decreased and the high-class imbalance problem is handled.

Unveiling relevant acoustic features for bird species automatic classification

The alarming rates of biodiversity loss has draw attention of the scientific community in developing quantitative characterization methods of the state of life variety, ecosystem and habitats. One way to address such aspects involves acoustically monitoring the bird population, given its relation with the health and status of the ecosystem. Automated computational tools can be adopted to enhance the processing and interpretation of birdsongs. The performance and reliability of the system heavily depends on the input features extracted from the acoustic recordings. Numerous acoustical features for characterizing birdsongs are reported in the literature, however the determination of the most relevant ones remains elusive. Moreover, literature evidences a marked focus on classification or detection performance, providing limited details on the set of features that contribute to high performance. This study investigates the problem of discerning relevant acoustical features when dealing with the automatic classification of eight Colombian bird species. We adopt different audio signal processing techniques, namely temporal, spectral, cepstral and chroma analyses, to construct a heterogeneous set of features. Feature selection is implemented using principal components analysis, ReliefF feature ranking, and genetic algorithms. By considering nearest neighbors classifiers, support vector machines, neural networks and bayesian classifiers, 49 distinct machine learning models are tested. Selection schemes are fine-tuned looking for maximizing the classification performance. Our results demonstrate that effective machine learning classification of bird species can be achieved by refining a heterogeneous set of features through feature selection. Consistently, our work identifies Mel frequency Cepstral coefficients, spectral decrease, and spectral rolloff point as the most recurrent features across various feature selection schemes. By incorporating these features in a heterogenous subset totaling a minimum of 19 features, classification performances above 95% can be achieved using a nearest neighbor classifier with the Manhattan distance. These findings are particularly notable as they are derived from processing field-noise corrupted acoustic recordings. The outcomes of this study yield potential acoustical features suitable for the characterization of birdsongs. In addition, our methodological design may serve as a reference for extending the assessment of new features and machine learning models, thereby contributing to the development of effective and feasible conservation tools.

An Intelligent Attack Detection Framework for the Internet of Autonomous Vehicles with Imbalanced Car Hacking Data

The modern Internet of Autonomous Vehicles (IoVs) has enabled the development of autonomous vehicles that can interact with each other and their surroundings, facilitating real-time data exchange and communication between vehicles, infrastructure, and the external environment. The lack of security procedures in vehicular networks and Controller Area Network (CAN) protocol leaves vehicles exposed to intrusions. One common attack type is the message injection attack, which inserts fake messages into original Electronic Control Units (ECUs) to trick them or create failures. Therefore, this paper tackles the pressing issue of cyber-attack detection in modern IoV systems, where the increasing connectivity of vehicles to the external world and each other creates a vast attack surface. The vulnerability of in-vehicle networks, particularly the CAN protocol, makes them susceptible to attacks such as message injection, which can have severe consequences. To address this, we propose an intelligent Intrusion detection system (IDS) to detect a wide range of threats utilizing machine learning techniques. However, a significant challenge lies in the inherent imbalance of car-hacking datasets, which can lead to misclassification of attack types. To overcome this, we employ various imbalanced pre-processing techniques, including NearMiss, Random over-sampling (ROS), and TomLinks, to pre-process and handle imbalanced data. Then, various Machine Learning (ML) techniques, including Logistic Regression (LR), Linear Discriminant Analysis (LDA), Naive Bayes (NB), and K-Nearest Neighbors (k-NN), are employed in detecting and predicting attack types on balanced data. We evaluate the performance and efficacy of these techniques using a comprehensive set of evaluation metrics, including accuracy, precision, F1_Score, and recall. This demonstrates how well the suggested IDS detects cyberattacks in external and intra-vehicle vehicular networks using unbalanced data on vehicle hacking. Using k-NN with various resampling techniques, the results show that the proposed system achieves 100% detection rates in testing on the Car-Hacking dataset in comparison with existing work, demonstrating the effectiveness of our approach in protecting modern vehicle systems from advanced threats.

PERBANDINGAN ANALISIS REGRESI LOGISTIK BINER DAN NAÏVE BAYES CLASSIFIER UNTUK MEMPREDIKSI FAKTOR RESIKO DIABETES

Diabetes is a global health problem that is increasing in prevalence worldwide. This study compares the performance of two data analysis methods, namely binary logistic regression and naïve bayes classifier in predicting diabetes risk. This study aims to identify factors that significantly affect diabetes risk and classify diabetes risk using binary logistic regression, then compare the classification with the naive bayes classifier algorithm. Binary logistic regression models the relationship between independent predictor variables and binary dependent variables, while naïve bayes classifier uses the assumption of independence between variables. In this study, both methods were evaluated based on accuracy, sensitivity, specificity and positive predictive value. The results show that the factors that influence the risk of diabetes are Age, Gender, Polyuria, Polydipsia, Genital thrush, Itching, Irritability, and Partial paresis. Furthermore, the binary logistic regression results have a higher classification accuracy (92.31%) compared to the naïve bayes classifier (84.61%). Therefore, binary logistic regression was identified as the best method to predict diabetes risk in the context of this study

Sentiment Analysis of the Indriver Online Ojek Application using the Naïve Bayes Classifier Method

According to statista.com, there are 73.1 million online motorcycle taxi users in Indonesia and there are 68.1 million active online motorcycle taxi users in Indonesia especially in the province of North Sumatra, there are 43,811 online motorcycle taxi drivers. The Indriver online motorcycle taxi application is an international online transportation service that gives passengers and drivers the freedom to negotiate prices. Sentiment analysis analizes text to determine positive, negative, or neutral sentiments. The method commonly used in sentiment analysis is the Naive Bayes Classifier method. This research uses quantitative methods to analyze sentiment toward the InDriver online motorcycle taxi application by utilizing the Naïve Bayes Classifier algorithm. User review data is collected from reviews on the Google Play Store, then cleaned and converted into a format suitable for statistical analysis. To analyze sentiment towards the InDriver online motorcycle taxi application using the Naïve Bayes Classifier method, collecting review data and user comments using the Python library and the Visual Studio code application, carrying out preprocessing, TF-IDF weighting, dividing the data into 70% and 30%, after that conducting testing using naïve Bayes classifier algorithm, as well as carrying out evaluation using a confusion matrix. The results of calculating the level of accuracy using the Naïve Bayes method for sentiment classification can be said to be good, this can be seen from the accuracy results on a dataset of 1393 with a comparison of training data and test data of 7:3, obtaining an accuracy value of 76%, precision of 71% , recall of 81% and f1-score of 76%. The results of this research analysis produced superior positive sentiment totaling 677 and negative sentiment totaling 608 while neutral was 90

Performance Evaluation of Predicting IoT Malicious Nodes Using Machine Learning Classification Algorithms

The prediction of malicious nodes in Internet of Things (IoT) networks is crucial for enhancing network security. Malicious nodes can significantly impact network performance across various scenarios. Machine learning (ML) classification algorithms provide binary outcomes ("yes" or "no") to accurately identify these nodes. This study implements various classifier algorithms to address the problem of malicious node classification, using the “SensorNetGuard” dataset. The dataset, comprising 10,000 records with 21 features, was preprocessed and used to train multiple ML models, including Logistic Regression, Decision Tree, Naive Bayes, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). Performance evaluation of these models followed the ML workflow, utilizing Python libraries such as scikit-learn, Seaborn, Matplotlib, and Pandas. The results indicated that the Naive Bayes classifier outperformed others with an accuracy of 98.1%. This paper demonstrates the effectiveness of ML classifiers in detecting malicious nodes in IoT networks, providing a robust predictive model for real-time application. The “SensorNetGuard” dataset is available on the IEEE data port and Kaggle platform.