Search

A photoplethysmography method has recently been widely used to noninvasively measure blood volume changes during a cardiac cycle. Photoplethysmogram (PPG) signals are sensitive to artifacts that negatively impact the accuracy of many important measurements. In this paper, we propose an efficient system for detecting PPG signal artifacts acquired from a fingertip in the public healthcare database named Multiparameter Intelligent Monitoring in Intensive Care (MIMIC) by using 11 features as the input of the random forest algorithm and classified the signals into two classes: acceptable and anomalous. A real-time algorithm is proposed to identify artifacts by using the method. The efficient Fisher score feature selection algorithm was used to order and select 11 relevant features from 19 available features that represented the PPG signal very effectively. Six machine learning algorithms (random forest, decision tree, Gaussian naïve Bayes, linear support vector machine, artificial neural network, and probabilistic neural network) were applied with the extracted features, and their classification accuracy was measured. Among them, the random forest had the best performance using only 11 out of 19 features with an accuracy of 85.68%. Our proposed method also achieved good sensitivity and specificity value of 86.57% and 85.09%, respectively. The proposed real-time algorithm can be an easy and convenient way for real-time PPG signal artifact detection using smartphones and wearable devices.

Penelitian ini bertujuan melakukan implementasi Probabilistic neural network dan Word Embedding dalam kasus sentiment analysis tentang tanggapan masyarakat tentang pemberian vaksin sinovac yangg diunggah di Twitter dan 3 class:positif, negative dan netral. Metode yang dipilih adalah metode klasifikasi Probabilistic Neural Network. Sebelum melakukan klasifikasi, praprocessing pada penelitian ini meliputi tokenizasi, normalisasi, menghilangkan emoticon, Convert Negasi, Stemming, Stopword Removal serta Word embedding. dataset yang digunakan berjumlah 1177 dataset dengan pembagiannya yaitu 560 dataset positif, 355 dataset negative dan 262 dataset netral. Program dirancang menggunakan Bahasa pemrograman python dengan beberapa library seperti keras, tensorflow dan pandas. Akurasi yang didapatkan pada pelatihan menggunakan Probabilistic Neural Network sebesar 91%. Hasil pengujian adalah penelitian ini mampu melakukan sentiment analysis dengan kesalahan sebesar 9%.

Nowadays, phishing attacks have become one of the major security threats that acquire the personal credentials of Internet users via forged websites for committing fraudulent financial transactions. The traditional phishing detection approaches employ single classification method in which the accuracy is more dependent on specific classification algorithm. A particular classifier may well perform on some dataset and less accurately on others. Hence, the framework for combining the complementary information of different classifiers is required to increase the prediction accuracy. This study assesses the performance of various neural network algorithms for selecting the base classifiers and models an ensemble method for detecting phishing websites. Based on the experimental results, Radial Basis Function (RBF), Generalized Radial Basis Function (GRBF), Probabilistic Neural Network (PNN), and Heteroscedastic Probabilistic Neural Network (HPNN) have been chosen as base classifiers for the proposed ensemble method. The proposed approach is focused on improving the performance of base classifiers individually as well as collaboratively for detecting phishing websites. Our proposed ensemble approach, Deep Ensemble Evidential Neural Network (DeepEEviNNet) is obtained by combining the outcome of base classifiers based on their weights for making the final decision. The optimal weight of each classifier is determined by the distance existing between the fusion result that is calculated using Dempster Shafer Theory (DST) and the ground truth. In addition, a novel categorical clustering algorithm named WEighted Fuzzy condense K-Modes (WEFKM) clustering is proposed to determine the RBF centers and Gaussian kernels of the base classifiers. The performance of DeepEEviNNet has been evaluated on various phishing datasets. The results obtained from the experiments reveal that DeepEEviNNet outperforms the stand-alone classification techniques as well as other ensemble methods for detecting phishing attacks.

PurposeBreast cancer is an important medical disorder, which is not a single disease but a cluster more than 200 different serious medical complications.Design/methodology/approachThe new artificial bee colony (ABC) implementation has been applied to probabilistic neural network (PNN) for training and testing purpose to classify the breast cancer data set.FindingsThe new ABC algorithm along with PNN has been successfully applied to breast cancers data set for prediction purpose with minimum iteration consuming.Originality/valueThe new implementation of ABC along PNN can be easily applied to times series problems for accurate prediction or classification.

To avoid the limitations of the assumptions for the traditional probabilistic seismic demand model (PSDM), this study proposed a novel PSDM for conducting fragility analysis of subway stations base...

Terrain identification is essential to autonomous control algorithm development for mobile robots. This article proposes an integrated framework to identify terrain parameters based on inertial, and driving current signals. Multiple sources are combined to reduce the instability caused by single signals. A dynamic model of the track-soil system is established as the theoretical basis of identification. All signals are processed in the time, frequency, and time-frequency domains. The features of each domain are generated by statistical methods. To analyze, and select superior feature categories, a maximum-relevance, and minimum-redundancy criterion based on Pearson's correlation is proposed to evaluate the priority of features. A probabilistic neural network is used to identify the category of terrain. All results are analyzed with two factors, source, and input, to find the most effective rule of the proposed framework. The crossing combination analysis is taken into consideration to explore all potential improvement. The results show that the driving current yields comparative identification accuracy as inertial signals. Compared to the single signal source, the method using the combined signal source can effectively improve the accuracy of terrain identification.

The principal intention of this work is to compare the performance of the supervised brain tumour segmentation methods. These segmentation methods are based on machine learning. First, the input MR brain image is denoised by employing the adaptive bilateral filter, and the image contrast is enhanced employing the histogram equalization. Then we retrieve the features from the pre-processed image. Among several feature extraction methods, this work uses the shape, intensity, and texture feature extractors. Subsequent to removing these three types of features, fragment the tumor dependent on these recovered segments. The supervised segmentation approach is used for this. Among several supervised segmentation methods, this work uses three machine learning methods, namely Probabilistic Neural Network (PNN), Artificial Neural Network (ANN), and Convolution Neural Network (CNN). Finally, the retrieved features are feed into these machine learning methods to segment the brain tumour regions. To find out the best machine learning approach, the performance of these three supervised machines learning methods is evaluated by four performance metrics. Based on these evaluations, the best segmentation approach is discovered. Four execution boundaries are utilized, in particular, Dice Similarity Coefficient (DSC), Positive Predictive Value (PPV), Jaccard list (JI), and Sensitivity (SEN) to analyze the presentation of the AI strategy. The experimental outputs exposed that the CNN makes greater than other methods.

Presently, the field of Internet of Things (loT) has been employed in diverse applications like Smart Grid, Surveillance, Smart homes, and so on. Precision Agriculture is a concept of farm management which makes use of IoT and networking concepts to improve the crop health. Recognition of diseases from the plant images is an active research topic which makes use of machine learning (ML) approaches. This paper introduces an effective rice plant disease identification and classification model to identify the type of disease from infected rice plants. The proposed method aims to detect three rice plant diseases such as Bacterial leaf blight, Brown spot, and Leaf smut. The proposed method involves a set of different processes namely image acquisition, preprocessing, segmentation, feature extraction and classification. At the earlier stage, IoT devices will be used to capture the image and stores it with a cloud server, which executes the classification process. In the cloud, the rice plant images under preprocessing to improvise the quality of the image. Then, fuzzy c-means (FCM) clustering method is utilized for the segmentation of disease portion from a leaf image. Afterwards, feature extraction takes place under three kinds namely color, shape, and texture. Finally, probabilistic neural network (PNN) is applied for multi-class classification. A detailed experimental analysis ensured the effective classification performance of the proposed method under all the test images applied.

Contemporary Smart Power Systems (SPNs) depend on Cyber-Physical Systems (CPSs) to connect physical devices and control tools. Developing a robust privacy-conserving intrusion detection method involves network and physical data regarding the setups, such as Supervisory Control and Data Acquisition (SCADA), for defending real data and recognizing cyber-attacks. A key issue in the implementation of SPNs is the security against cyber-attacks, targeting to interrupt SCADA operations and violate data privacy over the usage of penetration and data poisoning attacks. In this paper, a privacy-conserving framework, so-called PC-IDS, is proposed for realizing the privacy and safety features of SPNs through hybrid machine learning approach. The framework includes two key components. Primarily, a data pre-processing component is proposed for cleaning and transforming actual data into a different layout that accomplishes the aim of privacy conservation. Then, an intrusion detection component is proposed using a particle swarm optimization-based probabilistic neural network for the identification and recognition of malicious events. The performance of PC-IDS framework is evaluated by means of two commonly available datasets, i.e. the Power System and UNSW-NB15 datasets. The experimental outcomes highlight that the framework can proficiently protect data of SPNs and determine anomalous behaviours compared to numerous recent compelling state-of-the-art methods with respect to false positive rate (FPR), detection rate (DR) and computational processing time (CPT) by achieving 96.03% of DR, 0.18% FPR for Power System dataset and 95.91% of DR, 0.14% FPR for UNSW-NB15 dataset.

Coronavirus disease (COVID-19) has infected over more than 28.3 million people around the globe and killed 913K people worldwide as on 11 September 2020. With this pandemic, to combat the spreading of COVID-19, effective testing methodologies and immediate medical treatments are much required. Chest X-rays are the widely available modalities for immediate diagnosis of COVID-19. Hence, automation of detection of COVID-19 from chest X-ray images using machine learning approaches is of greater demand. A model for detecting COVID-19 from chest X-ray images is proposed in this paper. A novel concept of cluster-based one-shot learning is introduced in this work. The introduced concept has an advantage of learning from a few samples against learning from many samples in case of deep leaning architectures. The proposed model is a multi-class classification model as it classifies images of four classes, viz., pneumonia bacterial, pneumonia virus, normal, and COVID-19. The proposed model is based on ensemble of Generalized Regression Neural Network (GRNN) and Probabilistic Neural Network (PNN) classifiers at decision level. The effectiveness of the proposed model has been demonstrated through extensive experimentation on a publicly available dataset consisting of 306 images. The proposed cluster-based one-shot learning has been found to be more effective on GRNN and PNN ensembled model to distinguish COVID-19 images from that of the other three classes. It has also been experimentally observed that the model has a superior performance over contemporary deep learning architectures. The concept of one-shot cluster-based learning is being first of its kind in literature, expected to open up several new dimensions in the field of machine learning which require further researching for various applications.

The objective of this work is to select a Machine Learning Technique (MLT) to develop a fault diagnosis scheme for the power switching devices of the High Frequency link (HF link) in a grid-tied Photovoltaic (PV) system, without increasing the total number of sensors, and being capable to operate online. Artificial Neural Network (ANN), Support Vector Machine (SVM), Probabilistic Neural Network (PNN) and Naive Bayes (NB) algorithms are considered to solve the problem of fault classification. These four MLTs are compared using the specificity and sensitivity indexes. The inputs of the models are obtained from the mean value of the signals given by the Discrete Wavelet Transform (DWT) of the dc link voltage and the power extracted from the PV panels. Support vector machine algorithm is chosen as the most suitable classifier to diagnose single and simultaneous open circuit faults with lower computational effort. Simulation and real-time hardware-based experimental tests demonstrate that the MLTs are suitable and reliable to diagnose open circuit faults in a wide range of irradiance levels, ranging from 200 W/m2 to 1000 W/m2, even under 6 % and 12 % measurement errors, without increasing the overall system cost.

Physical and cognitive impairments decline the ability of elderly in execution of daily activities, such as eating, sleeping or taking medication. The proposed approach recognizes the activities performed in a smart home, and separates the normal from the anomalous activities. Moreover, we identify the anomalous days based on the number of activities performed in a day. We perform activity recognition by applying probabilistic neural network on the pre-segmented activity-data obtained from the sensors deployed at different locations in a smart home. We use H2O autoencoder to identify the anomalous from the normal instances of activities. We further categorize the anomalies based on the criteria such as missing or extra subevents, and unusual duration of activity. Since the ground truth of the anomalies is unavailable, we generate the ground truth using the boxplots of the duration, and the number of subevents in an activity. We provide the quantified results of activity recognition and anomaly detection that can be further used by the research community. A comprehensive evaluation of the proposed approach on two publicly available CASAS smart home datasets demonstrates its ability in the activity recognition and the correct identification of anomalies.

Evaluating the quality of environments within greenhouses has been an important aspect of greenhouse agriculture. However, the existing greenhouse monitoring systems are generally unable to evaluate the quality of greenhouse environments, and at present, it is evaluated manually, which leads to the failure of timely and effective evaluation. Therefore, how to evaluate the greenhouse environment quality accurately, quickly, and automatically is an urgent issue to be tackled. Inspired by the optimization algorithm, clustering algorithm, and probabilistic neural network (PNN), this article proposes the application of a novel PNN evaluation algorithm to a decentralized greenhouse monitoring system to evaluate the environment quality in real time. First, the original training samples are clustered by an improved $K$ -means clustering algorithm (called $K$ -means- $\alpha $ ), and the training samples near the cluster centroids are used as new ones to establish the PNN structure. Then, the particle swarm optimization (PSO) algorithm (whereby the fitness function is defined as the PNN’s classification error rate on the test samples) is used to iteratively optimize smoothing factors adopted by different classes of pattern units in the PNN. The optimized PNN is finally obtained (called $\alpha $ -PSO-M-PNN). The experimental results demonstrate that compared with conventional PNN-based algorithms, the optimized PNN algorithm has the advantages of a simpler network structure, higher classification accuracy, and requiring fewer training samples. In addition, it is more suitable for the microprocessor-based monitoring system because its computational and storage requirements are within the limits of the microprocessor.

In the paper, the flight time deviation of Lithuania airports has been analyzed. The supervised machine learning model has been implemented to predict the interval of time delay deviation of new flights. The analysis has been made using seven algorithms: probabilistic neural network, multilayer perceptron, decision trees, random forest, tree ensemble, gradient boosted trees, and support vector machines. To find the best parameters which give the highest accuracy for each algorithm, the grid search has been used. To evaluate the quality of each algorithm, the five measures have been calculated: sensitivity/recall, precision, specificity, F-measure, and accuracy. All experimental investigation has been made using the newly collected dataset from Lithuania airports and weather information on departure/landing time. The departure flights and arrival flights have been investigated separately. To balance the dataset, the SMOTE technique is used. The research results showed that the highest accuracy is obtained using the tree model classifiers and the best algorithm of this type to predict is gradient boosted trees.

Quick increase in web and system advancements has prompted significant increase in number of attacks and intrusions. Identification and prevention of these attacks has turned into an important part of security. Intrusion detection framework is one of the vital approaches to accomplish high security in computer systems and used to oppose attacks. Intrusion detection frameworks have reviled of dimensionality which tends to build time complexity and reduce resource use. Therefore, it is desirable that critical components of information must be examined by interruption detection framework to decrease dimensionality. These reduced features are then fed to a HFFPNN for training and testing on NSL-KDD dataset. HFFPNN is the hybridization of feed forward neural network (FFNN) and probabilistic neural network (PNN). Pre-processing of NSL-KDD dataset has been done to convert string attributes into numeric attributes before training. Comparisons with recent and relevant approaches are also tabled. Experimental results show the prominence of HFFPNN technique over the existing techniques in terms of intrusion detection classification. Therefore, the scope of this study has been expanded to encompass hybrid classifiers.

Nowadays, digital businesses with diverse deployment models such as cloud, mobile and edge devices for the internet of things will impact traffic, in both volume and dynamicity, at unprecedented rates. Moreover, due to the recent advances in optical networks and systems, the complexity of provisioning lightpaths is growing dramatically. Hence, optical network operators are forced to change their insight and move toward intent-based and self-driven networking, to cost-efficiently accommodate these challenging requirements. In this regard, knowledge-defined networking (KDN) promises to play a paramount role in realizing flexible and self-driven optical networks. In this work, we focus on one of the key aspects in this environment, i.e., prediction of quality of service for unestablished lightpaths. KDN is a solution that introduces machine learning techniques into the control plane of the network, to cope with inevitable complexities that arise in enabling network to operate autonomously and faster. For this, five machine learning models are evaluated for the classification and regression approaches. Multilayer perceptron, radial basis function and generalized regression neural network (GRNN) models are used for both of the regression and classification approaches, while support-vector machine and probabilistic neural network (PNN) models are used only for the classification scenario. Also, to discard the redundant features (among the considered experimental features) in the classification approach, input features are selected using the analysis of variance (ANOVA) test. The proposed models can accelerate and handle a significant part of operations in the closed-loop architecture of knowledge-defined optical networks, as a paradigm for designing self-driven optical networks. The best accuracies of quality of transmission prediction (classification approach) and optical signal-to-noise ratio estimation (regression approach) are achieved using PNN (with average accuracy of 99.6 ± 0.5%) and GRNN (with R-squared value of 0.957), respectively.

Speech Emotion Recognition (SER) has been widely used in many fields, such as smart home assistants commonly found in the market. Smart home assistants that could detect the user’s emotion would improve the communication between a user and the assistant enabling the assistant to offer more productive feedback. Thus, the aim of this work is to analyze emotional states in speech and propose a suitable algorithm considering performance verses complexity for deployment in smart home devices. The four emotional speech sets were selected from the Berlin Emotional Database (EMO-DB) as experimental data, 26 MFCC features were extracted from each type of emotional speech to identify the emotions of happiness, anger, sadness and neutrality. Then, speaker-independent experiments for our Speech emotion Recognition (SER) were conducted by using the Back Propagation Neural Network (BPNN), Extreme Learning Machine (ELM), Probabilistic Neural Network (PNN) and Support Vector Machine (SVM). Synthesizing the recognition accuracy and processing time, this work shows that the performance of SVM was the best among the four methods as a good candidate to be deployed for SER in smart home devices. SVM achieved an overall accuracy of 92.4% while offering low computational requirements when training and testing. We conclude that the MFCC features and the SVM classification models used in speaker-independent experiments are highly effective in the automatic prediction of emotion.

In the recent years, the volume of text documents in the form of digital way has grown up extremely in size. As significance, there is a need to be competent to automatically bring together and classify the documents based on their content. The main goal of text classification is to partition the unstructured set of documents into their respective categories based on its content. The main aim of this research work is to automatically classify the documents which are stored in the personal computer into their relevant categories. This work has two significant phases. In the first phase, the important features are selected for classification and the second phase is the classification of text documents. For selecting the optimal features, this research work proposes a new algorithm, optimization technique for feature selection (OTFS) algorithm. To estimate the proficiency of proposed feature selection algorithm, the OTFS algorithm was compared with the existing approaches artificial bee colony, firefly algorithm, ant colony optimization and particle swarm optimization. In the second phase, this research work proposed machine learning-based automatic text classification (MLearn-ATC) algorithm for text classification. In classification, the MLearn-ATC algorithm was compared with widely used classification techniques probabilistic neural network, support vector machine, K-nearest neighbor and Naive Bayes. From this, the output of first phase is used as the input for classification phase. The decisive results establish that the proposed algorithms achieve the better accuracy for optimizing the features and classifying the text documents based on their content.

The proposed work illustrates a simple research approach to identify the type of fault in a three-phase overhead single-end-fed long transmission line. Multivariate statistical methods like principal component analysis (PCA) alone, and in combination with probabilistic neural network (PNN), have been applied here to classify fault. An attempt has been made to use the PCA features obtained from the analysis of electrical parameters for each of the faults, in two ways. The first approach of fault classification is based on analyzing the PCA features by a modified ratio-based analysis. In the second method, an attempt has been made to use the PCA features directly to a structured PNN model. Electromagnetic Transient Program simulation software has been used to simulate a transmission line model. Sending-end three-phase line currents corresponding to various faults carried out at different geometric distances along the transmission line have been analyzed in MATLAB environment. The proposed algorithms are tested with unknown and intermediate distant faults with variable fault resistance to validate the same. Finally, a comparative analysis of the proposed two methods is illustrated, which shows 100% classifier accuracy of both the models.

This paper comprehensively adopts acoustic emission monitoring signals, ground stress monitoring signals, mining production data, energy evolution process data, microseismic statistics, mine geological structure and multidisciplinary data in subjects such as engineering mechanics, as well as existing cognitive laws to study and establish radial basis function neural network model for time-varying process signal analysis. Based on focal region localization and time-space environment correction alignment, the study bases itself on probability statistics theory and big data analysis technology. It studies sample process characteristics and the governing law in the already microseismic and periodic weighting events to investigate the statistical laws, trends and critical characteristics behind the event sample data so that microseismic magnitude and risk degree can be predicted. By changing the parameters of the nonlinear transformation function of the Neuron to realize the nonlinear mapping and the linearization of the connection weight adjustment, the learning speed of the network is improved. Compared with other dynamic neural network models which can deal with time-varying signal classification, the computational complexity is greatly reduced.

In deep geological drilling processes, the geological environment becomes more complex with the increasing of the drilling depth; consequently, the risks of downhole incidents get higher. If not discovered in time, these downhole incidents may develop to serious drilling accidents, causing significant financial and environmental losses. In this paper, a new method is proposed to diagnose downhole incidents by extracting trend features in multi-time scales and establishing a probabilistic neural network based diagnosis model. There are two major contributions: First, a feature extraction method is proposed to produce trend features from original process signals in different time scales; Second, an incident diagnosis method based on a broad probabilistic neural network is proposed to achieve better diagnosis performance in an expanded input space. Industrial case studies are presented to demonstrate the effectiveness and practicability of the proposed method. Results show that the proposed method has superior performance in diagnosing downhole incidents for geological drilling processes.

Due to limitations imposed by cored wells, lithological data are often incomplete, and correct identification of lithofacies is problematic. Identification is actually an issue of pattern recognition, and based on newly proved findings, LightGBM (light gradient boosting machine) is considered to be an excellent pattern recognizer and, therefore, well suited for recognizing lithofacies. To remove remaining disadvantageous features and to further enhance the prediction performance of LightGBM, CRBM (continuous restricted Boltzmann machine) and AFSA (artificial fish swarm algorithm) are adopted as assistants to provide, respectively, high-quality learning data and to create optimal hyper-parameter settings during data processing. Subsequently, a predictor characterized by new ensemble learning is proposed, named CRBM-AFSA-LightGBM. To establish comprehensive verification, several validations are designed based on logging data derived from pre-salt carbonate reservoirs of the Santos Basin. Validations demonstrate the effectiveness and significance of integrating CRBM and AFSA; a further two validations are aimed at revealing whether a change in the learning data has an impact on prediction. To highlight the validation effect, PNN (probabilistic neural network) and SVM (support vector machine) are introduced as contrasting predictors. The test results demonstrate three important points: (1) CRBM and AFSA are preferred to assist in the capability of LightGBM; (2) the LightGBM-cored predictor performs better when compared with PNN-cored and SVM-cored predictors, especially when dealing with larger-scale learning data; (3) better robustness of the new predictor because reliable identification still can be achieved even when learning samples are sparse. Because all the validation evidences are optimistic, CRBM-AFSA-LightGBM is verified as a highly efficient and robust prediction tool for lithofacies identification in carbonate reservoirs.

As far as, air quality index (AQI) is concerned, the long duration lockdown that was applied in India in year 2020 due to Covid-19 pandemic was very fruitful. The reason being, due to complete ban on the movement of people and automobiles, the air became so pure and clean, and AQI value went much down. The secondary air pollution data of the lockdown duration, for Uttarakhand, is the base of this research work. This work attempts to design unsupervised and supervised classification models to classify the provided data into two classes i.e class 1 (‘clean’) and class 2 (‘hazardous’) using MATLAB. The techniques used are FCM clustering and Probabilistic neural network (PNN). Eventually, a comparative study of the performance of both models is performed.

The advancement made in the field of information technology paves way for developing and utilization of software based technique in various field. The application of software related technique in the field of agriculture was at its peak nowadays. The key problems faced in agriculture are decrease in quality and yield of crops. This was caused due to various reasons such as environmental condition, insects, pest and diseases caused through microorganism. Among these various reasons the major fall in quality and yield of crop was achieved due to variety of disease. So, disease prediction at earlier stage in crops is considered as significant. Many methods were developed for prediction of disease in crop. But, effective prediction with better accuracy was not achieved in any of the developed method. Therefore for solving this issue the proposed system is developed. In the proposed method Whale Optimization Based Artificial Neural Network (WOANN) is used for prediction of different kinds of disease in tomato. Prior to classification of disease in tomato the process of colour image segmentation is carried out using FA in order to attain effective disease prediction ultimately. Further to evaluate the performance of the proposed system some statistical parameter such as accuracy, specificity, sensitivity, error rate, F1 score are calculated. And to prove that the proposed method functions effectively in disease prediction of tomato when compared to other conventional method such as Probabilistic Neural Network (PNN), K-Nearest Neighbour (K-NN) and Back Propagation Artificial Neural Network (BPANN) a comparison analysis is performed.

Classification is achieved through the categorisation of objects into predefined categories or classes, where the categories or classes are created based on a similar set of attributes of the object. This is referred to as supervised learning. Numerous methodologies have been formulated by researchers in order to solve classification problems effectively. These methodologies exhibit an uncomplicated structure and fast training, and are based on artificial intelligence, such as the probabilistic neural network (PNN). In this study, techniques to improve the accurateness of the PNN in solving classification problems have been analysed with the help of the water cycle algorithm (WCA), which is a population-based metaheuristic that imitates the water cycle in the real world. In the recommended solution, near-optimal solutions are created in order to regulate the arbitrary parameter selection of the PNN. In this study, it has also been suggested that the enhanced WCA (E-WCA) can be used to attain a balance between exploitation and exploration, so that premature conjunction and immobility of the population can be avoided. With the help of 11 standard benchmark datasets, the recommended solutions were verified. The results of the experiment substantiated that the WCA and E-WCA are capable of improving the weight parameters of the PNN, thereby imparting improved performance with respect to convergence speed and classification accuracy, compared with the initial PNN classifier.

Traffic stream determining is an essential part of the intelligent transportation management system. Precise prediction of traffic flow provides a basis for other tasks, like forecasting travel time. While traditional methods have some merits for improving traffic prediction precision in some ways, high precision, considering different circumstances, is still difficult to achieve. This article presents a short-term traffic flow prediction model based on the Modified Elman Recurrent Neural Network model (GA-MENN) to deal with this practical problem. In GA-MENN, the algorithm of Elman Recurrent Neural Network is modified, optimized through the Genetic Algorithm (GA) and considered weather conditions, weekday, hour and day's classification to forecast the vehicle velocity in Tehran streets and highways. The traffic data were collected from the online Google Map API service for 139 routs in 7 districts in Tehran. The method improves prediction precision and also lowers the prediction error rate, according to experimental results. Exploratory outcomes verify the superior performance of the proposed traffic condition prediction model over Regression Multi-layer Perceptron, Linear Regression, Logistic Regression, Probabilistic Neural Network, Regression Generalized Feedforward, Time-lag Recurrent Network, Support Vector Machine model, Elman neural network, K- NN model, ARIMA, Kalman filter model, Convolutional Neural Networks (CNNs), SARIMA, and Long Short-Term Memory (LSTM) model. To the best of our knowledge, this is the first occasion when that traffic stream is gauged in urban roads and avenues in this specific way.

This paper focuses on the real-time online monitoring and diagnosis framework for the angular misalignment of the robot spot-welding system, which can result in significant quality degradation of a weld nugget such as porosity. The data-driven approach is applied by installing the voltage and current sensors, collecting the associated mass data and processing them under normal and abnormal (angular misalignment) conditions. Two categories of features are extracted from the dynamic resistance (DR) and the voltage and current ones that are decomposed by wavelet transform (WT). The DR features are extracted from the DR profile and some critical features are selected by a t-test methodology. In the case of the WT-based features, the critical ones are selected by a max-relevance and min-redundancy (mRMR) and a sequential backward selection (SBS) wrapper. Consequently, three types of critical feature sets, such as DR features, WT features, and hybrid features combining those, are prepared to train machine learning-based models. Support vector machine (SVM) and probabilistic neural network (PNN) are applied to establish the diagnosis models, and the diagnostic accuracy and robustness are evaluated. Finally, the software for the on-line monitoring and diagnosis for angular misalignment of robot spot-welding system is developed and demonstrates its real-time applicability in an industrial site.

With the popularity of the Internet, people's lives are becoming more and more convenient. However, the network security problems are becoming increasingly serious. This paper, aiming to better protect users’ network security from the internal and external malicious attacks, briefly introduces the probabilistic neural network and principal component analysis method, and combines them for detection of network intrusion data. Simulation analysis of Probabilistic Neural Network (PNN) and Principal Component Analysis-Probabilistic Neural Network (PCA-PNN) are carried out in MATLAB software. The results suggest that the Principal Component Analysis (PCA) algorithm greatly reduce the dimension of the original data and the amount of calculation. Compared with PNN, PCA-PNN has higher accuracy and precision rate, lower false alarm rate, and faster detecting speed. Moreover, PCA-PNN has better detecting performance when there are few training samples. In summary, PCA-PNN can be used for the detection of network intrusion threat.

Detecting accidents is of great importance since they often impose significant delay and inconvenience to road users. This study compares the performance of two popular machine learning models, Support Vector Machine (SVM) and Probabilistic Neural Network (PNN), to detect the occurrence of accidents on the Eisenhower expressway in Chicago. Accordingly, since the detection of accidents should be as rapid as possible, seven models are trained and tested for each machine learning technique, using traffic condition data from 1 to 7 min after the actual occurrence. The main sources of data used in this study consist of weather condition, accident, and loop detector data. Furthermore, to overcome the problem of imbalanced data (i.e., underrepresentation of accidents in the dataset), the Synthetic Minority Oversampling TEchnique (SMOTE) is used. The results show that although SVM achieves overall higher accuracy, PNN outperforms SVM regarding the Detection Rate (DR) (i.e., percentage of correct accident detections). In addition, while both models perform best at 5 min after the occurrence of accidents, models trained at 3 or 4 min after the occurrence of an accident detect accidents more rapidly while performing reasonably well. Lastly, a sensitivity analysis of PNN for Time-To-Detection (TTD) reveals that the speed difference between upstream and downstream of accidents location is particularly significant to detect the occurrence of accidents.

Automatic plant species identification is one of the recent and fascinating research area as plants are crucial element of ecosystem. Several plant species exist with significant importance but most of us are unaware of the diversity of plant species available on earth. Their utility to humans starts as oxygen provider, food source, and medicinal compounds essential for medicines that are difficult to develop in right proportions. Being the first living habitants of earth, they have roots far deeper in the ecosystem than any living being. Hence, it is utmost important to develop automatic plant species identification system in which the digital image of the plant is given as input and the label of the plant is determined by the system. In this paper, we have focused on three different aspects (i) Significance of threshold (ii) Feature descriptor that can best describe the leaf images and (iii) Proposed a novel classification method called Multi class Twin Support Vector Machine which in an extension of widely used Twin Support Vector Machine classifier. The performance of the proposed method is compared with SVM, Multi Birth SVM and Probabilistic Neural Network. It is observed that the proposed classifier outperforms all the aforementioned classifiers on publicly available datasets.

Healthcare performs a key role in the health of humans in the world. While gathering a huge amount of medical data, the problems will appear on the classification of healthcare data. In this work, a fuzzy hybridised convolutional neural network (FCNN) model is stated to guess the class of healthcare data. This model collects the information from the data set and builds the decision table based on the collected features from data sets. The attributes that are unrelated are deleted by using principal component analysis algorithm. The decision of normal and cardiac disease is described by using FCNN classifier. Using the data sets from UCI (University of California Irvine) repository the estimation of the presented model is carried on. The performance of the authors’ classification technique is measured by various metrics such as accuracy, F-measure, G-mean, precision, and recall. The experimental results while compared with some of the existing machine learning methods such as probabilistic neural network, support vector machine and neural network, show the higher performance of FCNN. This model presented in this work acts as a decision support pattern in healthcare for therapeutic specialists.

The existence and persistence of financial statement fraud (FSF) are detrimental to the financial health of global capital markets. A number of detective and predictive methods have been used to prevent, detect, and correct FSF, but their practicability has always been a big challenge for researchers and auditors, as they do not address real-world problems. In this paper, both supervised and unsupervised approaches are employed for analysing the financial data obtained from China’s stock market in detecting FSF. The variables used in this paper are 18 financial datasets, representing a fraud triangle. Additionally, this study examined the properties of five widely used supervised approaches, namely, multi-layer feed forward neural network (MFFNN), probabilistic neural network (PNN), support vector machine (SVM), multinomial log-linear model (MLM), and discriminant analysis (DA), applied in different real-life situations. The empirical results show that MFFNN yields the best classification results in detection of fraudulent data presented in financial statement. The outcomes of this study can be applied to different types of financial statement datasets, as they present a practical way for constructing predictive models using a combination of supervised and unsupervised approaches.

The structure of water jet cutting machine is complex and the components are closely related. Its fault characteristics often have the characteristics of nonlinearity, coupling, uncertainty and complex causality. Traditional fault diagnosis methods have been difficult to solve the problem of water jet cutting machine fault detection quickly and effectively. As a new talent in the field of intelligent fault diagnosis, machine learning can independently mine the representative diagnostic information hidden in the original data and directly establish the accurate mapping relationship between the original data and the operating state, which has been increasingly applied in industrial diagnosis. In this paper, the application of intelligent fault diagnosis method for water jet cutting machine is explored. Data acquisition system of water jet cutting machine is built. Aiming at several common faults of water jet cutting machine, a fault diagnosis model is established based on PNN network, and the network is trained and tested with actual collected data. The results show that the probabilistic neural network model can better realize the fault diagnosis of common faults of water jet cutting machine.

Conventional predictive Artificial Neural Networks (ANNs) commonly employ deterministic weight matrices, therefore, their prediction is a point estimate. Such a deterministic nature in ANNs causes the limitations of using ANNs for medical diagnosis, law problems, and portfolio management in which not only discovering the prediction but also the uncertainty of the prediction is essentially required. In order to address such a problem, we propose a predictive probabilistic neural network model, which corresponds to a different manner of using the generator in the conditional Generative Adversarial Network (cGAN) that has been routinely used for conditional sample generation. By reversing the input and output of ordinary cGAN, the model can be successfully used as a predictive model, moreover, the model is robust against noises since adversarial training is employed. In addition, to measure the uncertainty of predictions, we introduce the entropy and relative entropy for regression problems and classification problems, respectively. The proposed framework is applied to stock market data and an image classification task. As a result, the proposed framework shows superior estimation performance, especially on noisy data, moreover, it is demonstrated that the proposed framework can properly estimate the uncertainty of predictions.

The problem of people with diabetes causes a condition known as Diabetic Retinopathy (DR). It is most common in the elderly age peoples. As diabetes progresses, patients' perceptions may begin to deteriorate and cause DR. People lose their sight because of this disease. To deal with DR, early detection is required. Patients will have to be examined by doctors regularly which is a waste of time and energy. In recent time, number of DR cases are increasing exponentially due the modern stress life style. Therefore, it is necessary to automate the detection and diagnosis process of the DR. In this paper, machine learning (ML) techniques are used to diagnose early DR. These are the Probabilistic Neural Network (PNN), Support Vector Machine (SVM), Bayesian Separation and K-Means Clustering. These methods will be tested and compared with choosing the best method. 3000 images being processed for training and testing. Features are extracted from these fundus images using image processing techniques. After research, the results confirm that SVM is the best way to detect DR early with a higher level of accuracy compared to other classifiers.

Road users classification plays an important role in the transportation management. With lower price of LiDAR, road user classification based on roadside Light Detection and Ranging (LiDAR) data has been a new approach in the transportation field. It is also essential for other intelligent transportation technology. In this paper, a method developed for road user classifications was proposed with roadside LiDAR data. The proposed method can be divided into four parts: background filtering, point clustering, feature selection, and road user classification. Five features of road users were selected based on the characteristic of road users’ point clouds. A comprehensive database was established and the classification performance of five machine learning methods including random forest, support vector machine, Probabilistic neural network, back propagation neural network, and AdaBoost were evaluated by F1score and F1macro. The results presented that AdaBoost had the best classification performance than others (The value of F1macro was 0.642).

Plants play a crucial role in nature and the well-being of the population. They have a significant contribution towards ecological stability and are also sources of our needs like food, medicine, and essential commercial products. As a result of massive scale deforestation, topsoil erosion, and habitat destruction, both the number and type of plants' existing species are steadily declining. So, plantation and identification and classification of plant species are essential for preserving plant species and accelerated farm as it will help in the better understanding of plants. Nevertheless, they are difficult to exercise as plant identification needs domain knowledge and experience. However, due to advances in machine learning and deep learning, this problem is tackled correctly. Various machine Learning and Deep Learning algorithms like Support Vector Machine, Artificial Neural Network, Convolutional Neural Network, Probabilistic Neural Network have successfully experimented on plant leaf images to identify the species with near correct accuracy. This article attempts a comparative analysis of various approaches used for plant identification. Several experiments with Swedish leaves confirm the effectiveness of machine learning and CNN based classification model.

In this chapter, we have introduced in details the miniTUBA system, and how to apply the miniTUBA dynamic Bayesian network (DBN) approach to analyze a typical use case in the areas of host-pathogen interactions using high throughput microarray data. The DBNs are powerful to model the stochastic evoluation of a set of random variables over time. Since the biological processes and various measurement errors are stochastic in nature, DBN has been considered as a suitable technique to study biological networks and pathways. Bayesian networks (BNs) and DBNs are based on a multinomial distribution. This distribution is very flexible, and each node has a different parameterization. Therefore, it is very feasible to use DBNs to model the dynamics of biological systems and responses to parameter perturbations. However, although a few applications for both Bayesian network and DBNs to modeling gene expression data have been discussed and reported, their usefullness remains to be shown with more well-understood pathways (Xia, et al., 2004). Programmed cell death (i.e., apoptosis) pathways are well studied and important for all plant and animal organisms. We first demonstrated in this report how the DBN analysis can be used to predict crucial genes for a cell death pathway, which led to correct experimental verification. Two major challenges in DBN analysis for biological network modeling exist. First, continuous gene expression data has to be descretized, leading to the loss of information. The descretization simplifies the computation and stablizes the predicated results. However, current equal quantile and interval descretization methods do not often reflect the biological realities. The customized descretization method is too time consuming and may not correlate with the unknown truth either. Therefore, alternative approaches will need to be explored to improve the descretization and minimize loss of information. How to find reliable ways to model continuous data remains to be a major challenge in the DBN and other modeling studies. Second, it is a big challenge to identify the correct time steps (i.e., Markov lags) for a DBN modeling. By default, we require all variables have the same time step size. However, it might be possible to allow a mixture of different time step sizes. The time scale likely differ between variables. To identify the relevant time scale, we may allow different discretization schemes. While more finely discretized variables offer slower changes, it might be difficult to determine how many are appropriate. The generation of very large sizes of discretizations is also time consuming. One solution is to allow mixtures of time steps in the learning step. However, it is in practice very difficult because the current step depends on a range of past experiences. If the previous time steps are not multiplies of each other, a complex splining function is usually needed to dynamically interpolate the missing data. Alternatively, we can explicitly search for an optimum informative time step. A DBN search will favor small time steps because it means more data to be used. However, if the data represents only more interpolated data, it would not help. While DBN analysis can be improved in different directions, the two areas of DBN research with the largest impact are probably the discretization and correct time step setting. Besides addressing the above challenges, dynamic Bayesian networks can further be improved through different directions: (i) those strong links (or edges) are conserved among top networks and can be detected by consensus analysis (Fig. 4). (ii) cross-species

Lithology prediction, especially for carbonate and volcanic reservoirs, is universally viewed as a crucial job in early petroleum exploration because the predicted lithology information is significant to some basic geological work such as stratigraphic correlation or sedimentation modeling. Although being capable to provide effective results in many practices, the typical methods normally cannot reach the goals when dealing with the complex lithology prediction. In order to solve such problem perfectly, a new hybrid neural network, a transformation of probabilistic neural network (PNN) by integrating CRBM (continuous restricted Boltzmann machine) and (PSO) particle swarm optimization, is proposed in this article. The raw data with high correlation can be filtered by CRBM, and then changes to be a new data set with low correlation. The utilization of PSO is to optimize the window length of probability density distribution for each kind of pattern. Then, with the help of CRBM and PSO, PNN becomes potential to make accurate judgments for test points. The data used for validation is collected from 8 cored wells of A oilfield. In two designed experiments, the proposed method gives higher prediction accuracies compared with other validated models, all of which are over 90%. Test results convincingly prove that the new hybrid network is capable to provide an effective solution for the complex lithology prediction, and the predicted results are reliable enough to serve as the reference data for further geological studies.

Since it is difficult for the traditional fault diagnosis method based on dissolved gas analysis (DGA) to meet today’s engineering needs in terms of diagnostic accuracy and stability, this paper proposes an artificial intelligence fault diagnosis method based on a probabilistic neural network (PNN) and bio-inspired optimizer. The PNN is used as the basic classifier of the fault diagnosis model, and the bio-inspired optimizer, improved salp swarm algorithm (ISSA), is used to optimize the hidden layer smoothing factor of PNN, which stably improves the classification performance of PNN. Compared with the traditional SSA, the sine cosine algorithm (SCA) and disruption operator are introduced in ISSA, which effectively improves the exploration capability and convergence speed. To verify the engineering applicability of the proposed method, the ISSA-PNN model was developed and tested using sensor data provided by Jiangxi Province Power Supply Company. In addition, the method is compared with machine learning methods such as support vector machine (SVM), back propagation neural network (BPNN), multi-layer perceptron (MLP), and traditional fault diagnosis methods such as the international electrotechnical commission (IEC) ratio method. The results show that the proposed method has a strong learning ability for complex fault data and has advantages in accuracy and robustness compared to other methods.

This study aimed to establish a machine learning (ML)-based rice blast predicting model to decrease the appreciable losses based on short-term environment data. The average, highest and lowest air temperature, average relative humidity, soil temperature and solar energy were selected for model development. The developed multilayer perceptron (MLP), support vector machine (SVM), Elman recurrent neural network (Elman RNN) and probabilistic neural network (PNN) were evaluated by F-measures. Finally, a sensitivity analysis (SA) was conducted for the factor importance assessment. The study result shows that the PNN performed best with the F-measure (β = 2) of 96.8%. The SA was conducted in the PNN model resulting in the main effect period is 10 days before the rice blast happened. The key factors found are minimum air temperature, followed by solar energy and equaled sensitivity of average relative humidity, maximum air temperature and soil temperature. The temperature phase lag in air and soil may cause a lower dew point and suitable for rice blast pathogens growth. Through this study’s results, rice blast warnings can be issued 10 days in advance, increasing the response time for farmers preparing related preventive measures, further reducing the losses caused by rice blast.

Smart toothbrushes equipped with inertial sensors are emerging as high-tech oral health products in personalized health care. The real-time signal processing of nine-axis inertial sensing and toothbrush posture recognition requires high computational resources. This paper proposes a recurrent probabilistic neural network (RPNN) for toothbrush posture recognition that demonstrates the advantages of low computational resources as a requirement, along with high recognition accuracy and efficiency. The RPNN model is trained for toothbrush posture recognition and brushing position and then monitors the correctness and integrity of the Bass Brushing Technique. Compared to conventional deep learning models, the recognition accuracy of RPNN is 99.08% in our experiments, which is 16.2% higher than that of the Convolutional Neural Network (CNN) and 21.21% higher than the Long Short-Term Memory (LSTM) model. The model we used can greatly reduce the computing power of hardware devices, and thus, our system can be used directly on smartphones.

Representing maritime traffic patterns and detecting anomalies from them are key to vessel monitoring and maritime situational awareness. We propose a novel approach -- referred to as GeoTrackNet -- for maritime anomaly detection from AIS data streams. Our model exploits state-of-the-art neural network schemes to learn a probabilistic representation of AIS tracks and a contrario detection to detect abnormal events. The neural network provides a new means to capture complex and heterogeneous patterns in vessels' behaviours, while the \textit{a contrario} detector takes into account the fact that the learnt distribution may be location-dependent. Experiments on a real AIS dataset comprising more than 4.2 million AIS messages demonstrate the relevance of the proposed method compared with state-of-the-art schemes., Comment: IEEE Transactions on Intelligent Transportation Systems

The development of medical images has facilitated the diagnosis of brain diseases. The diagnosis of brain medical images has the characteristics of uneven distribution of categories and different costs of misclassification. Therefore, traditional classification algorithms are used in clinically confirmed MRI brains. When a medical image is used as a training set to construct a classification model, the classification effect is poor and it is easy to be insensitive to the positive class, which makes it difficult for the brain disease auxiliary diagnosis system to have high accuracy and weak generalization ability. The research purpose of this paper is to study the assistant diagnosis system of brain diseases based on the uneven distribution of medical image categories. In order to improve the performance of the assistant diagnosis system of brain diseases, this paper designs a cost-sensitive probabilistic neural network CS-PNN brain by introducing cost-sensitive this system is an auxiliary diagnosis system for diseases, and the reliability of the system is verified by experiments. It can be known from experiments that the cost-sensitive probabilistic neural network CS-PNN assisted diagnosis system for brain diseases designed in this paper increases with the cost of positive misclassifications and negative misclassifications, and the classification accuracy rate of CS-PNN continues to increase. (01) = 4 achieves the best classification performance of 97%. The research in this article provides new ideas for solving the problems of uneven distribution of categories and misclassification costs in MRI brain medical images, so as to develop a brain disease auxiliary diagnosis system with stronger generalization ability, thereby improving the diagnosis of brain tumors. Accuracy and reduce missed diagnosis.

This paper is dedicated to analyzing the performance of an online classification algorithm called sequential heteroscedastic probabilistic neural network (SHPNN). The aforementioned algorithm is a variant of probabilistic neural networks (PNNs). This algorithm has the advantage of being structurally flexible to match the complexities of the data space. Another distinctive feature of this algorithm is the fact that it can achieve roughly the same level of accuracy compared to its counterparts while having an acceptable speed in training phase. But perhaps its most important quality is the fact that SHPNN’s structure contains valuable information about the underlying statistics of the data. In this paper the derivation of SHPNN formulation is presented and discussed. The following analysis includes comparison between the performance of SHPNN and other similar algorithms. Furthermore, the interpretability of this network is analyzed by visualizing the learning process.

Rice is the most significant cultivated harvests everywhere throughout the world, specifically in Asian nations. Nowadays the evaluation of rice quality has a great impact on the market due to adulteration by plastic rice and stones. The assessment of rice is prepared physically by experienced a rancher which is a tedious and monotonous errand and in particular it is a dangerous strategy where the rice might be pulverized by growth pollution. In this paper, a quick, programmed and non-destructive assessment practice is endeavored to measure the nature of rice based on deep learning neural system model. The pre-processing is started by the Median channel to expel noises from the input pictures. By utilizing Fuzzy c-means clustering, the edges of the rice pictures are appropriately depicted. The features like contour, edge, and region are chosen with the assistance of a genetic algorithm. A probabilistic neural system is created to characterize the portioned rice picture. The presentation of PNN model is introduced to show its viability with regards to exactness, accuracy, f-score, and review, and the outcomes are compared with the existing SVM method.

The paper explores the possibility of forecasting such dangerous meteorological phenomena as a thunderstorm by applying five types of neural network to the output data of a hydrodynamic model that simulates dynamic and microphysical processes in convective clouds. The ideas and the result delivered in [1] are developed and supplemented by the classification error calculations and by consideration of radial basic and probabilistic neural networks. The results show that forecast accuracy of all five networks reaches values of 90%. However, the radial basis function has the advantages of the highest accuracy along with the smallest classification error. Its simple structure and short training time make this type of neuralnetwork the best one in view of accuracy versus productivity relation.

Today, the entire world is suffering and fear with the epidemic of coronavirus. The statistics of coronavirus as on current data, 213 countries affected by this pandemic, more than 1.1 cr people, suffering from this killer virus, and about 6L fatality are recording. This virus is spreading speedily, and the patients are mainly suffering from breathing. The patient having previous health issues will get more possibility of this disease. In this work, try to evaluate the COVID 19 patient x-ray images by using DL (deep learning) techniques developed on the grouping of a recurrent neural network (RNN) and a correlational network to identify COVID-19 automatically is used in the prediction of high-risk outbreaks to learn on the prognostic code sequences of patients. By use of RNNs helps the model to assess difference in patient status in terms of time and thereby improve predictive precision. A correlational neural network is to identify the salient features for CORONAVIRUS, and these features are feed into a Probabilistic neural network (PNN) for better corona diagnosis. The experimental result gives improved accuracy for analyzing coronavirus disease.