Search

The growing Internet of Things (IoT) landscape requires robust security; traditional rule-based systems are insufficient, driving the integration of machine learning (ML) for effective intrusion detection. This paper provides an inclusive overview of research efforts focused on harnessing ML methodologies to fortify intrusion detection within IoT. Tailored feature extraction techniques are pivotal for achieving high detection accuracy while minimizing false positives. The study employs the IoT23 dataset from Kaggle and incorporates four optimization algorithms – Particle Swarm Optimizer, Whale-Pearson optimization algorithm, Harris-Hawks Optimizer, and Support Vector Machine with Particle Swarm optimization algorithm (SVM-PSO) – for feature extraction and selection. A comparison with ML algorithms such as logistic regression, decision tree and naïve Bayes classifier highlights Harris-Hawks Optimizer as the most effective. Furthermore, ensemble methods, particularly the fusion of random forest with HHO optimization, yield an impressive accuracy of 99.97%, surpassing AdaBoost and XGBoost approaches. This paper underscores the application of diverse ensemble learning techniques to enhance intrusion detection precision and efficiency within the intricate IoT landscape, effectively tackling the challenges posed by its complex and ever-changing nature. [ABSTRACT FROM AUTHOR]

The rapid expansion of information resources on the internet has made it increasingly challenging for users to get valuable and relevant information. Recommendation systems (RSs) play a vital role in addressing this issue by providing personalized suggestions based on user preferences. However, the open nature of RSs allows for the injection of fake profiles by malicious users, leading to biased ratings and manipulation of the overall system results. These fake profiles have been intended to promote or degrade specific items, which creates bias in the system and compromises the user experience. The motivation behind these fake profiles is to promote or degrade a particular item, which affects the system and makes it more biased toward certain items. This research paper aims to detect and eliminate such fake profiles in RSs, ensuring users receive genuine and reliable results. The proposed approach utilizes a voting ensemble (VE) method, combining the strengths of multiple classification algorithms including Quadratic Discriminant Analysis (QDA), K-Nearest Neighbor (KNN), and Naive Bayes classifier (NBC). By utilizing these classifiers, the system enhances its robustness against malicious activities. The proposed VE method effectively identifies the fake profiles with a minimum number of parameters, and it shows high detection accuracy on Movielens datasets. It has been demonstrated that the proposed method gives 99.4%, 99.5%, 99.4% and 99.6% accuracy using NBC, QDA, KNN, and VE, respectively and it outperforms all other competing methods in terms of accuracy, precision, recall and MCC. [ABSTRACT FROM AUTHOR]

Visual observation to recognize and identify aerial objects is a means to protect air sovereignty, and the delay can endanger it. Visual observation can be done through Ground-to-Air (GTA) or by approaching through Air-to-Air (ATA) using binoculars. The observation needs accuracy and speed to speed up the decision-making process. We propose an Artificial Intelligence (AI) system that receives voice input to recognize and identify an Unmanned Aerial Aircraft (UAA). We employed Naïve Bayes Classifier (NBC) that processes inputs from voice-to-text tools containing the observed UAA's feature. With 70 UAA samples, the AI system achieved an accuracy of 79% and a WER of 4.16%. [ABSTRACT FROM AUTHOR]

Traffic accidents are still the main cause of fatalities, injuries and significant delays in highways. Understanding the accident contributing factor is imperative to increase safety in a traffic network. Recent research confirms that predictive modeling is an important tool to comprehend accident contributing factors. However, little effort has been put forward to explain complex machine learning models and their feature effects in accident prediction models. Thus, this study aims to build predictive models based on different machine learning methods and tries to explain the most contributing factors by using Shapley values which was developed based on game theory. Decision Trees, Neural Networks with Multilayer Perceptron (MLP), Support Vector Classifier, Case-Based Reasoning and Naive Bayes Classifier were used to predict the injury severity in accidents. Belt usage, alcohol consumption and speed violations were found as the most effective features and MLP gave the highest accuracy among all the applied predictive models. [ABSTRACT FROM AUTHOR]

The COVID-19 virus, which was first confirmed in Wuhan, China, in 2019, has spread worldwide, and as of the beginning of 2022, more than 240 million cases and 4.77 million deaths have been confirmed. This present study aims to classify the significant risk of death for COVID-19 patients. Two statistical classification techniques, i.e., Bayesian logistic regression and Naive Bayes classifier, are implemented in this study. The performance of both methods is measured using several indicators. This study found that the significant explanatory variable involved in classifying the death risk for COVID-19 patients is Comorbidity, based on the output of both methods. Thus, more care is needed for those with comorbidities if they have been exposed to the virus. [ABSTRACT FROM AUTHOR]

This study, using the Naïve Bayes classifier, proposes a new descriptive model for conducting a comparative review analysis on the tourism domain. The proposed model seeks to improve the understanding of tourists' experiences by visually representing their opinions between various tourist destinations. The model predicts the sentiment of tourist reviews using the Naive Bayes classifier, which is then used to generate visualizations that facilitate comparative analysis. The comparison takes two tourism sites as experimental examples. The results, in terms of accuracy, show 80% and 64% for the first and the second site respectively. The study discusses the model's implementation, effectiveness, and potential applications in the tourism industry. The findings of the study suggest that the proposed model can provide valuable insights into tourist experiences and help tourism stakeholders, such as the managements of tourism sites, the future or soon coming tourists to make informed decisions based on comparative analysis. [ABSTRACT FROM AUTHOR]

In this approach, we considered classifying poems written in Marathi, one of the popular Indian languages, into nine categories. Using this classification, a person who is unaware of the Marathi Language can come to know what kind of emotion the poem shows. Here we considered tf-idf to represent the features of the poem. We have used Univariate feature selection(Chi2), Tree-based models, L1-based feature selection, and Recursive Feature Elimination to select top-ranked features. For nine categories of poems - Fear, Joy, Love(Prem), Sadness, Vir(Courage), Wonder, Anger, Depression, and Peace - the Naive Bayes classifier achieves maximum accuracy of 85% with Chi2 feature selection. Then we considered six categories of poems for classification. Experimentation is done using Naive Bayes and Neural Network machine learning algorithms. Among all feature selection methods, with Chi2 feature selection, the highest accuracy achieved is 97%. [ABSTRACT FROM AUTHOR]

As a unique type of social network, the online gaming industry is a fast-growing, changing, and men-dominated field which attracts diverse backgrounds. Being dominated by male users, game developers, game players, game investors, the non-inclusiveness and gender inequality reside as salient problems in the community. In the online gaming communities, most women players report toxic and offensive language or experiences of verbal abuse. Symbolic interactionists and feminists assume that words matter since the use of particular language and terms can dehumanize and harm particular groups such as women. Identifying and reporting the toxic behavior, sexism, and harassment that occur in online games is a critical need in preventing cyberbullying, and it will help gender diversity and equality grow in the online gaming industry. However, the research on this topic is still rare, except for some milestone studies. This paper aims to contribute to the theory and practice of sexist toxic language detection in the online gaming community, through the automatic detection and analysis of toxic comments in online games chat logs. We adopted the MaXQDA tool as a data visualization technique to reveal the most frequently used toxic words used against women in online gaming communities. We also applied the Naïve Bayes Classifier for text mining to classify if a chat log content is sexist and toxic. We also refined the text mining model Laplace estimator and re-tested the model's accuracy. The study also revealed that the accuracy of the Naïve Bayes Classifier did not change by the Laplace estimator. The findings of the study are expected to raise awareness about the use of genderbased toxic language in the online gaming community. Moreover, the proposed mining model can inspire similar research on practical tools to help moderate the use of sexist toxic language and disinfect these communities from gender-based discrimination and sexist bullying. [ABSTRACT FROM AUTHOR]

Objectives: the crop yield prediction rate has been improved using a machine learning regression algorithm (MLR) using a Naïve Bayes classifier. The optimal utilization of fertilizer is enhanced based on a potential of hydrogen (pH) value and alkalinity of soil. Method: the implementation of the proposed algorithm has been carried out by considering the various types of soils, percentage of nutrients like potassium (Pm), nitrogen (N) and Phosphorus (P) in the soil in that region, four to five years of data analysis of the amount of rainfall, atmosphere humidity, and crop yield to fertilizer utilization ratio of a particular region and duration. Finding: The designed system has a model to accurately and precisely predict crop yield and give required recommendations to the end-user regarding fertilizer ratio depending on soil and weather conditions of the land to improve the crop yield, thereby increasing the revenue of farmers. Novelty: the farmers have to determine the expected crop yield and required fertilizer by themselves. It is achieved more accurately by implementing the proposed algorithm compared to the existing Random Forest (RF) algorithm using data mining Decision support system. The analysis has been carried out on the above-mentioned attributes of data by adopting the data pre-processing, data testing, and validation to achieve a precise and accurate crop yield and fertilizer utilization model. [ABSTRACT FROM AUTHOR]

In this paper I propose a wrapped feature selection method using Recursive Feature Elimination and Cross-validated selection. In my work I use Bernoulli Naïve Bayes classifier on the NSL-KDD dataset. [ABSTRACT FROM AUTHOR]

Hazards and disasters have always negative impacts on the way of life. Landslide is an overwhelming natural as well as man-made disaster that causes loss of natural resources and human properties throughout the world. The present study aimed to assess and compare the prediction efficiency of different models in landslide susceptibility in the Kysuca river basin, Slovakia. In this regard, the fuzzy decision-making trial and evaluation laboratory combining with the analytic network process (FDEMATEL-ANP), Naïve Bayes (NB) classifier, and random forest (RF) classifier were considered. Initially, a landslide inventory map was produced with 2000 landslide and non-landslide points by randomly divided with a ratio of 70%:30% for training and testing, respectively. The geospatial database for assessing the landslide susceptibility was generated with the help of 16 landslide conditioning factors by allowing for topographical, hydrological, lithological, and land cover factors. The ReliefF method was considered for determining the significance of selected conditioning factors and inclusion in the model building. Consequently, the landslide susceptibility maps (LSMs) were generated using the FDEMATEL-ANP, Naïve Bayes (NB) classifier, and random forest (RF) classifier models. Finally, the area under curve (AUC) and different arithmetic evaluation were used for validating and comparing the results and models. The results revealed that random forest (RF) classifier is a promising and optimum model for landslide susceptibility in the study area with a very high value of area under curve (AUC = 0.954), lower value of mean absolute error (MAE = 0.1238) and root mean square error (RMSE = 0.2555), and higher value of Kappa index (K = 0.8435) and overall accuracy (OAC = 92.2%). Unlabelled Image • Landslide susceptibility modeling plays a significant role in disaster prevention. • Three models (FDEMATEL-ANP, NBC, and RFC) were applied and compared. • The ROC curve was used for evaluating and comparing the performance of the results. • Random forest classifier (RFC) produced the best result for susceptibility assessment. • An accurate model of susceptibility is helpful for landslide risk mitigation and planning. [ABSTRACT FROM AUTHOR]

Nowadays, the use of Open Data has become more common and prominent, but there are a lot of questions regarding its quality. Most of the revised researches deal with the quality of Open Data portals, rather than estimation of the open datasets quality. Therefore, the main idea of this research is lowering to the level of the dataset itself in order to assess how much such data is downloaded by end users of Open Data portals on the basis of general dataset characteristics. A model for predicting the number of downloads of open datasets based on their general characteristics was constructed using the Na've Bayes Classifier. Based on the obtained results, it is discussed if the certain dataset character is good predictor of open dataset downloading and to what extent. [ABSTRACT FROM AUTHOR]

On April 17, 2019, Indonesia will hold a presidential general election. Analysis of public sentiment towards presidential candidates and vice-presidential candidates of Indonesia in 2019 through social media especially Facebook, is needed to find out the popularity and public sentiments towards the candidates. Data was obtained through comments on Facebook posts of three major news media in Indonesia, namely Detik (@detikcom), TribunNews (@tribunnews), and Liputan6 (@liputan6online). The Opinion is classified using the data mining approach using the Naive Bayes Classifier algorithm. This research found that 40.52% of the comments were talking about presidential and vice-presidential candidate Joko Widodo and Maruf Amin (Jokowi-Maruf) meanwhile presidential candidate Prabowo Subianto and his running mate Sandiaga Uno (Prabowo-Sandi) dominate the comments by 59.48%. On the contrary, for the sentiment polarity results, the Jokowi-Maruf dominate by 56.76% positive sentiment and 43.24% negative sentiment, while the Prabowo-Sandi received 24.21% positive sentiment and 75.79% negative sentiment. [ABSTRACT FROM AUTHOR]

Hadith is believed to be the main source of Islam after Qur'an. The simplicity of obtaining hadith information is currently supported by global access using the internet. The abundance of hadith literature sometimes finds difficulties to obtain the information that needed. Therefore, information extraction is required to facilitate the searching of information in hadith. In this study, the name indexing in Indonesian translation of hadith from nine narrators was built. The model was built using Named Entity Recognition with Naïve Bayes classifier. The features used in this study are title case, POS tag and unigram. This study experimented with individual features and features that were combined. Precision, recall, and F1-Score are employed as evaluation metrics. F1-Score is used in this study to measure the performance of named entity and features. The results of experiments extracted 258 people's names from 13870 token data from 100 Indonesian hadith texts and show that implementing the combination of all features can achieve 82.63% of F1-Score. [ABSTRACT FROM AUTHOR]

Abstract The implementation of an efficient adaptive e-learning system requires the construction of an effective student model that represents the student's characteristics, among those characteristics, there is the learning style that refers to the way in which a student prefers to learn. Knowing learning styles helps adaptive E-learning systems to improve the learning process by providing customized materials to students. In this work, we have proposed an approach to identify the learning style automatically based on the existing learners' behaviors and using web usage mining techniques and machine learning algorithms. The web usage mining techniques were used to pre-process the log file extracted from the E-learning environment and capture the learners' sequences. The captured learners' sequences were given as an input to the K-modes clustering algorithm to group them into 16 learning style combinations based on the Felder and Silverman learning style model. Then the naive Bayes classifier was used to predict the learning style of a student in real time. To perform our approach, we used a real dataset extracted from an e-learning system's log file, and in order to evaluate the performance of the used classifier, the confusion matrix method was used. The obtained results demonstrate that our approach yields excellent results. [ABSTRACT FROM AUTHOR]

Abstract In this paper, we investigate a set of methods for textual Arabic Dialect Identification, where we considered word-level and sentence-level approaches. We used three classifiers, namely: Linear Support Vector Machine L-SVM, Bernoulli Naive Bayes BNB and Multinomial Naive Bayes MNB. Then we combined them by using a voting procedure. We carried out experiments on two sets of dialects: the first one, PADIC, which consists of parallel sentences in Maghrebi and Middle Eastern dialects; and the second, a set of Algerian dialects only, that we built manually. For the Arabic dialects, we obtained an average accuracy of 92%. For Algerian dialects, our approach yielded an average accuracy of about 76%. [ABSTRACT FROM AUTHOR]

Financial credit scoring is one of the most crucial processes in the finance industry sector to be able to assess the credit-worthiness of individuals and enterprises. Various statistics-based machine learning techniques have been employed for this task. “Curse of Dimensionality” is still a significant challenge in machine learning techniques. Some research has been carried out on Feature Selection (FS) using genetic algorithm as wrapper to improve the performance of credit scoring models. However, the challenge lies in finding an overall best method in credit scoring problems and improving the time-consuming process of feature selection. In this study, the credit scoring problem is investigated through feature selection to improve classification performance. This work proposes a novel approach to feature selection in credit scoring applications, called as Information Gain Directed Feature Selection algorithm (IGDFS), which performs the ranking of features based on information gain, propagates the top m features through the GA wrapper (GAW) algorithm using three classical machine learning algorithms of KNN, Naïve Bayes and Support Vector Machine (SVM) for credit scoring. The first stage of information gain guided feature selection can help reduce the computing complexity of GA wrapper, and the information gain of features selected with the IGDFS can indicate their importance to decision making. Regarding the classification accuracy, SVM accuracy is always better than KNN and NB for Baseline techniques, GAW and IGDFS. Also, we can conclude that the IGDFS achieved better performance than generic GAW, and GAW obtained better performance than the corresponding single classifiers (baseline) for almost all cases, except for the German Credit dataset, IGDFS + KNN has worse performance than generic GAW and the single classifier KNN. Removing features with low information gain could produce conflict with the original data structure for KNN, and thus affect the performance of IGDFS + KNN. Regarding the ROC performance, for the German Credit Dataset, the three classic machine learning algorithms, SVM, KNN and Naïve Bayes in the wrapper of IGDFS GA obtained almost the same performance. For the Australian credit dataset and the Taiwan Credit dataset, the IGDFS + Naive Bayes achieved the largest area under ROC curves. [ABSTRACT FROM AUTHOR]

Word Sense Disambiguation (WSD) is the process of identifying the correct sense of the word in the context. The most leading scheme used by WSD is machine learning approach, where a human expert provides examples of correctly disambiguated words, and a machine learning algorithm is used to induce a model from these examples. In this paper, Naive Bayes supervised classifier has been used to disambiguate words of Punjabi language. The feature extraction process plays a vital role in building the supervised machine learning models. For the proposed Punjabi WSD system, Bag of Words (BoW) and collocation models are used separately to extract relevant features. BoW model has used all words around target word while collocation model has used two words before and two words after the target word as features. Both the models have used a common training data set to build the model. It has been observed that the selection of smoothing parameter for Naive Bayes has a significant impact on its performance. This proposed work has been tested on 150 most ambiguous noun words selected form Punjabi WordNet having 6 or more senses. During the process of building the model, fine senses of ambiguous words have been merged to produce coarse sense on the basis of manual analysis of lexical relations of WordNet. The accuracy of the proposed system has been calculated independently for BoW and collocation model. The proposed WSD system achieves an accuracy of 89% for BoW model and 81% for collocation model. It has been concluded that BoW model performs better than the collocation model for WSD task for Punjabi language. [ABSTRACT FROM AUTHOR]

The output of traditional analysis of student feedback for class room delivery of faculty suffers from inaccuracy due to non consideration of the influence of various direct and indirect quality features related to student such as regularity in class attendance, effort, academic background, course outcomes achieved and positive attitude on the feedback measure. Consequently, the output of traditional faculty feedback analysis is not a true indicator of faculty effectiveness in the teaching learning process. The paper presents a proactive and outcome based faculty feedback analysis model which uses Naïve Bayes Classifier to cull out and classify the feedback provided by each student into valid or invalid categories on the basis of the relative effect of aforementioned quality features on the feedback measure. The above quality features are used to refine the feedback measure. The method attempts to address the imprecision to overcome the limitations of the traditional model. Consequently, the output of faculty feedback analysis results in a more refined and accurate Faculty Effectiveness Index. The Faculty Effectiveness Index is calculated as weighted average of only the valid feedback measures with the validity of feedback taken as the associated weight. The classifier takes into consideration the independent contribution of each of the features as well as the multiple evidences of their occurrences in the feedback provided by each student. The method also suggests a comprehensive feedback form comprising of two parts namely subjective feedback part for eliciting feedback in traditional manner and an outcome based feedback part to collect information on the aforesaid quality features related to student which influence the feedback measure. [ABSTRACT FROM AUTHOR]

The prediction of the dinoflagellate red tide forming Karenia selliformis is a relevant task to aid optimized management decisions in marine coastal water. The objective of the present study is to compare different modeling approaches for prediction of Karenia selliformis occurrences and blooms. A set of physical parameters (salinity, temperature and tide amplitude), meteorological constraints (evaporation, air temperature, insolation, rainfall, atmospheric pressure and humidity), sampling months and sampling sites are used. The model prediction included general linear model (GLM), Bayesian Network (BN) and the simplest BN type which is, Naive Bayes classifier (NB). The results showed that three models incriminated high salinity in Karenia selliformis blooms and the sampling sites, mainly Boughrara lagoon, in the occurrences. The BN performed better than linear models (NB and GLM) for both Karenia selliformis occurrences and blooms prediction. This later is related to the facts that BN considered the inter-independency between predictive variables and that the relationships between the variables and the outcome are often non-linear such us; the transition to bloom situations appeared to be triggered by a salinity threshold. This study is useful in the management of this ecosystem so as to use the best disposal options in the early prediction of the toxic blooms. [ABSTRACT FROM AUTHOR]

Card transaction fraud is prevalent anywhere. Even with current preventive measures like the Europay, MasterCard and Visa (EMV) chips, possible weaknesses or loopholes can be exploited by fraudsters. This paper explores Naïve Bayes Classifier and clustering algorithms to detect fraud in credit card transactions. Data on fraud labels and arrival times of transactions were simulated by the Markov Modulated Poisson Process. Amounts of transactions for genuine and fraud transactions were simulated based on two Gaussian distributions. Kinds of spenders and types of fraudsters serve as the bases for the parameters used in the simulation of the data. Using the simulated data, EM clustering algorithm with three different initializations and Kmeans were applied to cluster transaction amounts into high, medium and low. The Naïve Bayes classifier algorithm was then applied to classify the transactions as good or fraud for the simulated data of 9 types of fraudsters across all clustering algorithms. Simulations and analyses were done using R software. Results include comparisons of true positive rates, false positive rates, and detection accuracies among the nine types of fraudsters across all clustering algorithms. For 3 clusters, (high, medium, low transaction amounts), the Naïve Bayes Method with clustering algorithms resulted to an average of 76% true positive (TP) detection, 18% false positive (FP) detection, with an overall accuracy of 81%. The same averages of TP, FP, and overall accuracy were obtained using 2 clusters (high, and low). EM clustering algorithm generated TP, FP, and overall accuracy of 80%, 16%, and 83% respectively. [ABSTRACT FROM AUTHOR]

Rock burst is often induced by the superposition of static and dynamic loads that produces failure with a sudden and violent release of elastic energy accumulated in rock and coal masses during underground activities. Casualties, deformation of the supporting structures and damage of the equipment on site are some of its consequences, hence producing a need to study its prediction. A novel application of Bayesian networks (BNs) to predict rock burst is proposed in this paper. In order to analyze the influence of the network structure, several networks are constructed with five parameters: Tunnel depth (H), Maximum tangential stress of surrounding rock (MTS) (σ θ ), Uniaxial tensile strength of rock (UTS) (σ t ), Uniaxial compressive strength of rock (UCS) (σ c ) and Elastic energy index (W et ). The Expectation Maximization algorithm is employed to learn from a data set of 135 rock burst case histories with incomplete data, whereas belief updating is carried out by the Junction Tree algorithm. The model is validated with 8-fold cross-validation and with another new group of incomplete case histories that had not been employed during training of the BN, and the influence of the network structure on the classification results, as well as the advantages and limitations of different network structures, are discussed. Results suggest that BNs are able to satisfactorily deal with incomplete data, hence becoming a useful tool to predict the rock burst hazard at the initial stages of underground work. [ABSTRACT FROM AUTHOR]

In a large number of experimental problems, high dimensionality of the search area and economical constraints can severely limit the number of experimental points that can be tested. Within these constraints, classical optimization techniques perform poorly, in particular, when little a priori knowledge is available. In this work we investigate the possibility of combining approaches from statistical modeling and bio-inspired algorithms to effectively explore a huge search space, sampling only a limited number of experimental points. To this purpose, we introduce a novel approach, combining ant colony optimization (ACO) and naïve Bayes classifier (NBC) that is, the naïve Bayes ant colony optimization (NACO) procedure. We compare NACO with other similar approaches developing a simulation study. We then derive the NACO procedure with the goal to design artificial enzymes with no sequence homology to the extant one. Our final aim is to mimic the natural fold of 200 amino acids 1AGY serine esterase from Fusarium solani . [ABSTRACT FROM AUTHOR]

Parallel transmission lines are difficult to protect due to mutual coupling between circuits. This paper proposes a Naïve Bayes classifier (NBC) based fault detection and classification technique for protection of parallel transmission line involving inter-circuit faults. NBC is a good classification tool for larger data sets as the training process takes less time with greater accuracy. Input given to the fault detection module is the fundamental components of three phase current signals of both circuits. Input given to the fault phase identification and fault classification module is the fundamental component of three phase current signals and zero sequence currents of both the circuits. Seven separate classifiers are designed for fault phase identification for A1, B1, C1, A2, B2, C2 and G. From fault phase identification module faults are classified. Accuracy of the proposed method is 100% for fault detection and 99.99% for classification of fault from all the tested fault cases. Response time of the proposed method is within 10 ms for all the fault cases studied. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Information Science & Technology Association/Joho no Kagaku to Gijutsu is the property of Information Science & Technology Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This paper focuses on comparative study of calculation of classification error with classical PCA technique and our adaptive method. Frontal face image database with uniform lightening condition and color images with different orientations have taken in order to get better accuracy in the result. Conventional PCA algorithm is applied for dimensional reduction and calculating the classification error. The result generated from this technique is being compared with the result generated from our adaptive method Naïve Bayes Classifier. We have used Naïve Bayes Classifier for calculating the classification error of each feature vector instead of considering K largest Eigen-value as in PCA. A covariance matrix is arranged by considering the feature vector having the lowest K minimum error. We have applied the adaptive technique on 626 colored facial images with uniform illumination conditions and varying poses and 545 frontal facial images with uniform background to get the better accuracy. [ABSTRACT FROM AUTHOR]

Clustering and classification models, or hybrid models are the most widely used models that can handle the diverse nature of network intrusion detection systems (NIDS) dataset. Dirichlet process clustering technique is a non-parametric Bayesian mixture model that considers the data distribution of the dataset for the formation of distinct clusters. The number of clusters is not known a priori and it differs across different datasets. Determining the number of clusters based on the distribution of data instances can increase the performance of the model. Naive Bayes model, a supervised learning classification technique, maintains a better computational efficiency, by reducing the training time. In this paper, author proposes a hybrid model to exploit the positive aspect of proper clustering of data instances and the computational efficiency in building a NIDS. RIPPER algorithm is used to extract rules from the traffic description for updation of the rule database. Experiments were conducted in the KDD CUP'99 and SSENet-2011 datasets to study the performance of the proposed model. Also, a comparison of three hybrid methods with the proposed hybrid model was carried out. The results showed that the proposed hybrid model is superior in building a robust perimeter security device [ABSTRACT FROM AUTHOR]

Monte Carlo simulation is an effectivemethod for evaluating complex systems. Besides estimating the performancelevel of the system through Monte Carlo method, it is more wanted to identifykey factors in system operation so as to improve or redesign the system. When estimating the performance level, in order to obtainsufficientevaluation accuracy while keeping time cost as low as possible, its relation with confidence leveland number of simulation runs is explained according to probability and statistics theory. To identify key factors, a method ranking the significant influencing parameters automaticallyfor complex systemsbased on naive Bayes classifier (NBC) and kernel density estimator (KDE) is developed. NBC used for classification makes the method valid for all kinds of linear and nonlinear complex systems, and KDE contributes greatly to identifyingsignificant influencing parameters in automated manner.The method above is applied toa reusable launch vehicle (RLV) flight evaluation.Through the evaluation, bias of atmosphere density is identified as the most significant parameter which relies on the flight control mode in the terminal flight phase. [ABSTRACT FROM AUTHOR]

In the paper we describe a sequential classification scheme that iteratively explores levels of abstraction in the description of examples. These levels of abstraction represent attribute values of increasing precision. Specifically, we assume attribute values constitute an ontology (i.e., attribute value ontology) reflecting a domain-specific background knowledge, where more general values subsumes more precise ones. While there are approaches that consider levels of abstraction during learning, the novelty of our proposal consists in exploring levels of abstraction when classifying new examples. The described scheme is essential when tests that increase precision of example description are associated with costs – such a situation is often encountered in medical diagnosis. Experimental evaluation of the proposed classification scheme combined with ontological Bayes classifier (i.e., a näıve Bayes classifier expanded to handle attribute value ontologies) demonstrates that the classification accuracy obtained at higher levels of abstraction (i.e., more general description of classified examples) converges very quickly to the classification accuracy for classified examples represented precisely. This finding indicates we should be able to reduce the number of tests and thus limit their cost without deterioration of the prediction accuracy. [ABSTRACT FROM AUTHOR]

Naive Bayes algorithm captures the assumption that every attribute is independent from the rest of the attributes, given the state of the class attribute. In this study, we attempted to increase the prediction accuracy of the simple Bayes model by integrating global and local application of Naive Bayes classifier. We performed a large-scale comparison with other attempts that have tried to improve the accuracy of the Naive Bayes algorithm as well as other state-of-the-art algorithms on 28 standard benchmark datasets and the proposed method gave better accuracy in most cases. [ABSTRACT FROM AUTHOR]

The tremendous growth of the web-based applications has increased information security vulnerabilities over the Internet. Security administrators use Intrusion-Detection System (IDS) to monitor network traffic and host activities to detect attacks against hosts and network resources. In this paper IDS based on Naïve Bayes classifier is analyzed. The main objective is to enhance IDS performance through preparing the training data set allowing to detect malicious connections that exploit the http service. Results of application are demonstrated and discussed. In the training phase of the proposed IDS, at first a feature selection technique based on Naïve Bayes classifier is used, this technique identifies the most important HTTP traffic features that can be used to detect HTTP attacks. In the testing and running phases proposed IDS classifies the network traffic based on the requested service, then based on the selected features Naïve Bayes classifier is used to analyze the HTTP service based traffic and identifies the HTTP normal connections and attacks. The performance of the IDS is measured through experiments using NSL-KDD data set. The results show that the detection rate of the IDS is about 99%, the false-positive rate is about 1%, and the false-negative rate is about 0.25%; therefore, proposed IDS holds the highest detection rate and the lowest false alarm compared with other leading IDS. In addition, the proposed IDS based on Naïve Bayes is used to classify network connections as a normal or attack. And it holds a high detection rate and a low false alarm. [ABSTRACT FROM AUTHOR]

In this paper, we present a sequence classifier for detecting host intrusions from long process system call sequences. The proposed classifier (called SC2.2) is a naïve Bayes classifier that builds class conditional probabilities from Markov modeling of system call sequences. We describe the proposed classifier, and then provide experimental results on the widely used University of New Mexico's system call trace data sets. The results of our proposed classifier are benchmarked against leading classifiers, namely naive Bayes multinomial, C4.5 decision tree, RIPPER, support vector machine, and logistic regression. A key feature of the proposed classifier is its ability to handle efficiently arbitrarily long sequences, and zero transitional probabilities in the modeled Markov chain, by using a ?test, backtrack, scale and remultiply? technique, and using the m-estimate of conditional probabilities. Furthermore, it capitalizes on IEEE standard for floating-point arithmetic error specification to return classification confidence. Results show that the proposed classifier yields a better performance than standard classifiers on most datasets used. [ABSTRACT FROM AUTHOR]

• Parametric statistical tests to investigate balanced accuracy in discriminating pathological and normal voices. • Signal modeling and processing methods to obtain a suitable glottal waveform for each voice. • Extraction of six discriminating features through Kernel Principal Component Analysis (KPCA) • Modeling the posterior distribution of the balanced accuracy to calculate the p-value of the average of the cross-validation results. • Significant improved performance in discrimination of pathological voices. The ability to speak lucidly plays a key role in social relations. Consequently, the role of the larynx is quite important, and timely diagnosis of laryngeal diseases has proved to be crucial. In this study, a simple computational model for inverse of speech production model is employed to extract the glottal waveform using speech signal. This waveform has useful information about vocal folds performance in terms of providing evidence for distinguishing pathological disorders. Furthermore, obtaining the significance of classification results is important, because it leads to reliable inferences. This study utilizes the sustained vowel sound /a/ and a well-referenced database, namely MEEI. In this work, after extraction of six discriminating features by using appropriate signal modeling and processing methods and upon change of the feature space by using Kernel Principal Component Analysis (KPCA), a classifier consisting of Naïve Bayes and Fisher Linear Discriminant, is exerted on the feature sets. Regarding voice pathology detection, the proposed approach achieved a significant classification balanced accuracy 93.6 % ± 0.03 with p-value < 0.01 for normal and abnormal classification using the Beta distribution model for the posterior distribution of the average of the cross-validation results. The proposed features are also compared with some conventional features in this field. The results show significant improved performance for the proposed features in discriminating different types of pathological voices. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, we describe the design and evaluation of extractive summarization approach to assist the learners with reading difficulties. As existing summarization approaches inherently assign more weights to the important sentences, our approach predicts the summary sentences that are important as well as readable to the target audience with good accuracy. We used supervised machine learning technique for summary extraction of science and social subjects in the educational text. Various independent features from the existing literature for predicting important sentences and proposed learner dependent features for predicting readable sentences are extracted from texts and are used for automatic classification. We performed both extrinsic and intrinsic evaluation on this approach and the intrinsic evaluation is carried out using F-measure and readability analysis. The extrinsic evaluation comprises of learner feedback using likert scale and the effect of assistive summary on improving readability for learners’ with reading difficulty using ANOVA. The results show significant improvement in readability for the target audience using assistive summary. [Copyright &y& Elsevier]

Abstract: Due to the tremendous growth of network based services, intrusion detection has emerged as an important technique for network security. While variety of security techniques are being developed and a lot of research is going on intrusion detection, but the field lacks an integrated approach with high detection rate (recall) and precision for minority attacks namely R2L and U2R. However, the recall and precision goals are often conflicting and attacking them simultaneously may not work well, especially when some of the classes are rare. This paper presents a novel layered approach with multi-classifier by combining naïve bayes classifier (NBC) and naive bayes tree (NBTree) to improve detection rate and precision of minority class without hurting the performance of majority class. We identify important reduced feature set for each attack separately, to form layered approach. The proposed approach scales up the recall and precision for major as well as minor attacks, and keeps the false positives at acceptable level in intrusion detection. [Copyright &y& Elsevier]

Abstract: Intrusion detection is the process of monitoring and analyzing the events occurring in a computer system in order to detect signs of security problems. Today most of the intrusion detection approaches focused on the issues of feature selection or reduction, since some of the features are irrelevant and redundant which results lengthy detection process and degrades the performance of an intrusion detection system (IDS). The purpose of this study is to identify important reduced input features in building IDS that is computationally efficient and effective. For this we investigate the performance of three standard feature selection methods using Correlation-based Feature Selection, Information Gain and Gain Ratio. In this paper we propose method Feature Vitality Based Reduction Method, to identify important reduced input features. We apply one of the efficient classifier naive bayes on reduced datasets for intrusion detection. Empirical results show that selected reduced attributes give better performance to design IDS that is efficient and effective for network intrusion detection. [Copyright &y& Elsevier]

Abstract: This article presents the classification of blood characteristics by a C4.5 decision tree, a naïve Bayes classifier and a multilayer perceptron for thalassaemia screening. The aim is to classify eighteen classes of thalassaemia abnormality, which have a high prevalence in Thailand, and one control class by inspecting data characterised by a complete blood count (CBC) and haemoglobin typing. Two indices namely a haemoglobin concentration (HB) and a mean corpuscular volume (MCV) are the chosen CBC attributes. On the other hand, known types of haemoglobin from six ranges of retention time identified via high performance liquid chromatography (HPLC) are the chosen haemoglobin typing attributes. The stratified 10-fold cross-validation results indicate that the best classification performance with average accuracy of 93.23% (standard deviation=1.67%) and 92.60% (standard deviation=1.75%) is achieved when the naïve Bayes classifier and the multilayer perceptron are respectively applied to samples which have been pre-processed by attribute discretisation. The results also suggest that the HB attribute is redundant. Moreover, the achieved classification performance is significantly higher than that obtained using only haemoglobin typing attributes as classifier inputs. Subsequently, the naïve Bayes classifier and the multilayer perceptron are applied to an additional data set in a clinical trial which respectively results in accuracy of 99.39% and 99.71%. These results suggest that a combination of CBC and haemoglobin typing analysis with a naïve Bayes classifier or a multilayer perceptron is highly suitable for automatic thalassaemia screening. [Copyright &y& Elsevier]

Building intelligent course designing systems adaptable to the learners' needs is one of the key goals of research in e-learning. This goal is all the more crucial as gaining knowledge in an e-learning environment depends solely on computer medi- ated interaction within the learner group and among the learners and instructors. The patterns generated out of discussions among e-learners may reveal significant information regarding their learning growth. This paper presents an algorithm to predict the knowledge gain of students by analyzing their online interaction pattern amongst each other. This knowledge gain is used to categorize the students as prospective "gainers" or "non-gainers" through Naïve Bayes Classifier. The preferences of non-gainers in terms of instructor-oriented v/s peer-oriented interaction are subsequently obtained. The paper further suggests some of the remedial plans based on proven instructional strategies, to be adopted that may help learners strengthen their weak areas. The actual knowledge gain of learners is evaluated by performing paired t-test on the previous and post-test score pairs. [ABSTRACT FROM AUTHOR]

Abstract: Dictionary-based protein name recognition is often a first step in extracting information from biomedical documents because it can provide ID information on recognized terms. However, dictionary-based approaches present two fundamental difficulties: (1) false recognition mainly caused by short names; (2) low recall due to spelling variations. In this paper, we tackle the former problem using machine learning to filter out false positives and present two alternative methods for alleviating the latter problem of spelling variations. The first is achieved by using approximate string searching, and the second by expanding the dictionary with a probabilistic variant generator, which we propose in this paper. Experimental results using the GENIA corpus revealed that filtering using a naive Bayes classifier greatly improved precision with only a slight loss of recall, resulting in 10.8% improvement in F-measure, and dictionary expansion with the variant generator gave further 1.6% improvement and achieved an F-measure of 66.6%. [Copyright &y& Elsevier]

This paper presented an integrated method for unlabeled data assisted active learning in order to solve insufficient labeled training data in supervised learning. The proposed approach uses naive Bayes with expectation maximization (EM) to learn an initial query concept based on the labeled and formerly unlabeled images. Then the active learning algorithm selects the most useful images in database to query the user for labeling, and the query concept is updated iteratively according to user's feedbacks. Various experiments demonstrate that the proposed approach can improve the efficiency and accuracy of the learner and can grasp user's query concept quickly. [ABSTRACT FROM AUTHOR]

• A novel hybrid segmentation method is presented for sickle cell anaemia (SCA) detection. • The hybrid segmentation method consists of fuzzy C-means and four adaptive (local) thresholding techniques. • Performances of Naïve Bayes and KNN classifiers are studied and compared. • The classifiers used in the work provide superior SCA classification performance with the proposed hybrid segmentation method. Detection of anomalous cells by analyzing the microscopic blood smear plays a key role in identification of various blood diseases in medical field. It becomes easier with robust segmentation steps followed by a classifier for automatic recognition of normal as well as anomalous blood cells. Sickle cell anemia is a disease that changes shape of the red blood cell (RBC) into sickle cell which can be detected by use of automatic classifier. In this reported work, a novel robust hybrid segmentation method is introduced for segmenting normal RBCs and sickle cells. The hybrid segmentation process proposed here is a combination of two segmentation methods namely fuzzy C-means and adaptive (local) thresholding. Four different adaptive thresholding methods are experimented here in combination with fuzzy C-means. The features extracted for carrying out this analysis are major axis, minor axis, aspect ratio, area, perimeter, form factor (metric value), eccentricity and solidity of each of the cells present in the sample blood smear. These eight features are used for training and testing the classifiers. For classification, two supervised classifiers, i.e. Naïve Bayes classifier and K-nearest neighbor classifier have been demonstrated on a dataset of 10 image data samples and the evaluated results are compared for each of the hybrid combinations. The comparison shows that the proposed hybrid combination of fuzzy C-means and NICK's thresholding alongwith K-nearest neighbor classifier gives the best performance in terms of the considered performance parameters. This combination gives a significant classification accuracy of 98.87% with lowest computational time. [ABSTRACT FROM AUTHOR]

We consider the problem of learning a density function from observations of an unknown underlying model in a distributed setting, where the observations are partitioned into different sites. Applying commonly used density estimation methods such as Gaussian Mixture Model (GMM) or Kernel Density Estimation (KDE) to distributed data leads to an extensive amount of communication. A familiar approach to address this issue is to sample a small subset of data and collect them into a central node to run the density estimation algorithms on them. In this paper, we follow an alternative to the sub-sampling approach by proposing the nested Log-Poly model. This model provides an accurate density estimation from a small sized statistic of the entire data. In distributed settings, it transfers the small sized statistics from the client nodes to a central node. The estimation process is then run in the central node. The proposed model can be used in different learning tasks such as classification in supervised learning and clustering in unsupervised learning. However, the properties of nested Log-Poly make it a suitable model for one-dimensional density estimations in the distributed settings. This makes Log-Poly a good choice for naive Bayes classifier, where one-dimensional density estimation is required for every feature conditioned on the class label. We provide a theoretical analysis of the efficiency of our model in estimating a wide range of probability density functions. Our experiments show that nested Log-Poly outperforms the state of the art density estimators on several synthetic datasets. We compare the accuracy and the communication load of naive Bayes classifier using nested Log-Poly and other related density estimators on several real datasets. The experimental outcomes depict that nested Log-Poly has less communication load, while maintaining a competitive classification accuracy compared to similar methods that use the entire data. Moreover, we present a comprehensive comparison between nested Log-Poly and validated KDE with sub-sampling, in terms of the number of communicated variables and the number of bytes transferred between the clients and the central node. Nested Log-Poly provides comparable accuracy with the validated KDE with sub-sampling, while communicating fewer variables. However, our method needs to compute and transmit the variables with a high precision in order to accurately capture the details of the underlying distributions. • Introducing a new distributed learning algorithm for density estimation. • Reducing the communication costs. • Theoretical analysis for the efficiency of the proposed method. • Application for large-scale classification tasks. [ABSTRACT FROM AUTHOR]

• A voice pathology detection system grounded on simple and clear-cut foundations. • The features are designed precisely in terms of voice pathologies effects upon the speech signal. • The system delivers high accuracy with a low number of parameters. This paper proposes new features aiming to improve the performance of an automatic voice pathology detection system. The features are designed precisely in terms of voice pathologies effects upon the speech signal. The system is intended to deliver high accuracy with a low number of parameters. Kullback–Leibler divergence (KLD) applied to consecutive frames of the speech signal provides a measure of voice instability. In this work, the KLD is applied to frame's histogram and a modified form of its spectrum named higher amplitude suppression spectrum (HASS). The H-KLD (histogram KLD) and the HASS-KLD are two of the three features presently approached. An additional feature that provides the level of damping of the voice pitch period waveform is proposed, the short-term sample skewness of the signal. The H-KLD, the HASS-KLD, and the sample skewness are features employed along with mel-frequency cepstral coefficients (MFCC) in a voice pathology detection system. The system is composed of a Gaussian mixture models (GMM) classifier and two generalized extreme value (GEV) distribution classifiers. They are fused by means of a Gaussian naïve Bayes classifier. A standard subset of the Massachusetts Eye and Ear Infirmary (MEEI) voice disorders database is adopted for evaluating the system. The obtained global success rate of 99.55% shows that the proposed features are suitable for pathological voice quality assessment. [ABSTRACT FROM AUTHOR]

Background: The placenta is a maternal-fetal organ that develops during pregnancy and provides nutrients, oxygen, and removal of waste products to the growing fetus. Better understanding of the placenta promises to help improve health of mothers and children, given its influence on health lasting a lifetime. However, the placenta is poorly understood due to its variability across different species and no live functions available after the baby is delivered. The Placenta Atlas Tool (PAT) project aims to leverage advanced computational approaches to meld numerous and diverse datasets into an integrated resource to encourage a "systems biology" approach for study of both normal and abnormal placental development and function throughout gestation.Methods: In this study, we introduced a multi-layer framework to automatically identify PAT relevant research from PubMed and develop a Placenta Curated Research Dataset (PCRD) to ultimately support placenta research. This framework functions by multiple well-known Natural Language Processing (NLP) components; including Medical Subject Headings (MeSH) based Naïve Bayes classifier, abstract based text similarity comparison and MeSH based article prioritization to systematically filter out PAT relevant research publications for further data curation. In addition, we developed a user-friendly web application to incorporate human judgement at the final stage of publication identification.Results: We obtained 22,047 articles from PubMed, and programmatically identified 6086 articles that are highly relevant to PAT via our framework. To assess performance of the framework, we manually reviewed a random set of articles by using our web tool. Based on our review, accuracy of article classification is greater than 90% and accuracy of prioritization is greater than 80%.Conclusions: We developed a multi-layer publication identification framework to systematically identify PAT relevant publications from PubMed. This framework not only demonstrates good performance in identifying placenta related research, but also can be easily extended to support research in other scientific fields. [ABSTRACT FROM AUTHOR]