Search results

The purpose of this paper is to apply a multi‐objective Bayesian artificial immune system (MOBAIS) to feature selection in classification problems aiming at minimizing both the classification error and cardinality of the subset of features. The algorithm is able to perform a multimodal search maintaining population diversity and controlling automatically the population size according to the problem. In addition, it is capable of identifying and preserving building blocks (partial components of the whole solution) effectively.

The algorithm evolves candidate subsets of features by replacing the traditional mutation operator in immune‐inspired algorithms with a probabilistic model which represents the probability distribution of the promising solutions found so far. Then, the probabilistic model is used to generate new individuals. A Bayesian network is adopted as the probabilistic model due to its capability of capturing expressive interactions among the variables of the problem. In order to evaluate the proposal, it was applied to ten datasets and the results compared with those generated by state‐of‐the‐art algorithms.

The experiments demonstrate the effectiveness of the multi‐objective approach to feature selection. The algorithm found parsimonious subsets of features and the classifiers produced a significant improvement in the accuracy. In addition, the maintenance of building blocks avoids the disruption of partial solutions, leading to a quick convergence.

The originality of this paper relies on the proposal of a novel algorithm to multi‐objective feature selection.

This paper is concerned with the estimation of reservoir parameters in the presence of noise and outliers using neural network (NN) and Bayesian algorithm. The paper aims to discuss these issues.

Outlier detection is of great importance to prediction of time series data. A reliable predictive methodology is proposed based on NN and Bayesian algorithm to efficiency estimates of the parameters of a petroleum reservoir. This strategy is applied to estimate the parameters of Marun reservoir located in Ahwaz, Iran utilizing available geophysical well log data.

For an evaluation purpose, the performance and generalization capabilities of Bayes-ANN are compared with the common technique of back propagation (BP).

The experimental results demonstrate that the proposed hybrid Bayes-NN algorithm is able to reveal a better performance than conventional BP NN algorithms.

Helped oil and gas companies to estimation of petroleum reservoir parameters more accurate than other methods in the presence of noise and outliers.

Bayesian cubature (BC) has emerged to be one of most competitive approach for estimating the multi-dimensional integral especially when the integrand is expensive to evaluate, and alternative acquisition functions, such as the Posterior Variance Contribution (PVC) function, have been developed for adaptive experiment design of the integration points. However, those sequential design strategies also prevent BC from being implemented in a parallel scheme. Therefore, this paper aims at developing a parallelized adaptive BC method to further improve the computational efficiency.

By theoretically examining the multimodal behavior of the PVC function, it is concluded that the multiple local maxima all have important contribution to the integration accuracy as can be selected as design points, providing a practical way for parallelization of the adaptive BC. Inspired by the above finding, four multimodal optimization algorithms, including one newly developed in this work, are then introduced for finding multiple local maxima of the PVC function in one run, and further for parallel implementation of the adaptive BC.

The superiority of the parallel schemes and the performance of the four multimodal optimization algorithms are then demonstrated and compared with the k-means clustering method by using two numerical benchmarks and two engineering examples.

Multimodal behavior of acquisition function for BC is comprehensively investigated. All the local maxima of the acquisition function contribute to adaptive BC accuracy. Parallelization of adaptive BC is realized with four multimodal optimization methods.

The purpose of the food traceability is used to retain the good quality of raw material supply, diminish the loss and reduced system complexity.

The proposed hybrid algorithm is for food traceability to make accurate predictions and enhanced period data. The operation of the internet of things is addressed to track and trace the food quality to check the data acquired from manufacturers and consumers.

In order to survive with the existing financial circumstances and the development of global food supply chain, the authors propose efficient food traceability techniques using the internet of things and obtain a solution for data prediction.

The operation of the internet of things is addressed to track and trace the food quality to check the data acquired from manufacturers and consumers. The experimental analysis depicts that proposed algorithm has high accuracy rate, less execution time and error rate.

Machine scheduling plays an important role in most manufacturing industries and has received a great amount of attention from operation researchers. Production scheduling is concerned with the allocation of resources and the sequencing of tasks to produce goods and services. Dispatching rules help in the identification of efficient or optimized scheduling sequences. The purpose of this paper is to consider a data mining‐based approach to discover previously unknown priority dispatching rules for the single machine scheduling problem.

In this work, the supervised statistical data mining algorithm, namely Bayesian, is implemented for the single machine scheduling problem. Data mining techniques are used to find hidden patterns and rules through large amounts of structured or unstructured data. The constructed training set is analyzed using Bayesian method and an efficient production schedule is proposed for machine scheduling.

After integration of naive Bayesian classification, the proposed methodology suggests an optimized scheduling sequence.

This paper analyzes the progressive results of a supervised learning algorithm tested with the production data along with a few of the system attributes. The data are collected from the literature and converted into the training data set suitable for implementation. The supervised data mining algorithm has not previously been explored in production scheduling.

This paper seeks to develop an approach to problem localization and an algorithm to address the issue of determining the dependencies among system metrics for automated system management in ubiquitous computing systems.

This paper proposes an approach to problem localization for learning the knowledge of dynamic environment using probabilistic dependency analysis to automatically determine problems. This approach is based on Bayesian learning to describe a system as a hierarchical dependency network, determining root causes of problems via inductive and deductive inferences on the network. An algorithm of preprocessing is performed to create ordering parameters that have close relationships with problems.

The findings show that using ordering parameters as input of network learning, it reduces learning time and maintains accuracy in diverse domains especially in the case of including large number of parameters, hence improving efficiency and accuracy of problem localization.

An evaluation of the work is presented through performance measurements. Various comparisons and evaluations prove that the proposed approach is effective on problem localization and it can achieve significant cost savings.

This study contributes to research into the application of probabilistic dependency analysis in localizing the root cause of problems and predicting potential problems at run time after probabilities propagation throughout a network, particularly in relation to fault management in self‐managing systems.

The ultimate objective of the any e-Learning system is to meet the specific need of the online learners and provide them with various features to have efficacious learning experiences by understanding their complexities. Any e-Learning system could be much more improved by tracking students commitment and disengagement on that course, in turn, would allow system to have personalized involvements at appropriate times in order to re-engage learners. Motivations play a important role to get back the learners on the track could be done by analyzing of several attributes of the log files. This paper aims to analyze the multiple attributes which cause the learners to disengage from an online learning environment.

For this improvisation, Web based learning system is researched using data mining techniques in education. There are various attributes characterized for the disengagement prediction using web log file analysis. Though, there have been several attempts to include motivating characteristics in e-Learning systems are adapted, presently influence on cognition is acknowledged mostly.

Classification is one of the predictive data mining technique which makes prediction about values of data using known results found from different data sets. To find out the optimal solution for identifying disengaged learners in the online learning systems, Naive Bayesian (NB) classifier with Particle Swarm Optimization (PSO) algorithm is used which will classify the data set and then perform the independent analysis.

The experimental results shows that the use of unrelated variables in the class attributes will reduce the accuracy and reliability of a any classification model. However, the hybrid PSO algorithm is clearly more apt to find minor subsets of attributes than the PSO with NB classifier. The NB classifier combined with hybrid PSO feature selection method proves to be the best feature selection capability without degrading the classification accuracy. It is further proved to be an effective method for mining large structural data in much less computation time.

This paper has used the well-known machine learning (ML) computational algorithm with Internet of Things (IoT) devices to predict the COVID-19 disease and to analyze the peak rate of the disease in the world. ML is the best tool to analyze and predict the object in reasonable time with great level of accuracy. The Purpose of this paper is to develop a model to predict the coronavirus by considering majorly related symptoms, attributes and also to predict and analyze the peak rate of the disease.

COVID-19 or coronavirus disease threatens the human lives in various ways, which leads to deaths in most of the cases. It affects the respiratory organs slowly and this penetration leads to multiple organ failure, which causes death in some cases having poor immunity system. In recent times, it has drawn the international attention because of the pandemic threat that is harder to control the spreading of infection around the world.

This proposed model is implemented by support vector machine classifier and Bayesian network algorithm, which yields high accuracy. The K-means algorithm has been applied for clustering the data set models. For data collection, IoT devices and related sensors were used in the identified hotspots. The data sets were collected from the selected hotspots, which are placed on the regions selected by the government agencies. The proposed COVID-19 prediction models improve the accuracy of the prediction and peak accuracy ratio. This model is also tested with best, worst and average cases of data set to achieve the better prediction rate.

From that hotspots, the IoT devices were fixed and accessed through wireless sensors (802.11) to transfer the data to the authors’ database, which is dedicated in data collection server. The data set and the proposed model yield good results and perform well with expected accuracy rate in the analysis and monitoring of the recovery rate of COVID-19.

This study aims to introduce a methodology for optimal allocation of spinning reserves taking into account load, wind and solar generation by application of the univariate and bivariate parametric models, conventional intra and inter-zonal spinning reserve capacity as well as demand response through utilization of capacity outage probability tables and the equivalent assisting unit approach.

The method uses a novel approach to model wind power generation using the bivariate Farlie–Gumbel–Morgenstern probability density function (PDF). The study also uses the Bayesian network (BN) algorithm to perform the adjustment of spinning reserve allocation, based on the actual unit commitment of the previous hours.

The results show that the utilization of bivariate wind prediction model along with reserve allocation adjustment algorithm improve reliability of the power grid by 2.66% and reduce the total system operating costs by 1.12%.

The method uses a novel approach to model wind power generation using the bivariate Farlie–Gumbel–Morgenstern PDF. The study also uses the BN algorithm to perform the adjustment of spinning reserve allocation, based on the actual unit commitment of the previous hours.