Search

Nonoccurring behavior (NOB) studies have attracted the growing attention of scholars as a crucial part of behavioral science. As an effective method to discover both NOB and occurring behaviors (OB), negative sequential pattern (NSP) mining is successfully used in analyzing medical treatment and abnormal behavior patterns. At this time, NSP mining is still an active and challenging research domain. Most of the algorithms are inefficient in practice. Briefly, the key weaknesses of NSP mining are: 1) an inefficient positive sequential pattern (PSP) mining process, 2) a strict constraint of negative containment, and 3) the lack of an effective Negative Sequential Candidate (NSC) generation method. To address these weaknesses, we propose a highly efficient algorithm with improved techniques, named sc-NSP, to mine NSP efficiently. We first propose an improved PrefixSpan algorithm in the PSP mining process, which connects to a bitmap storage structure instead of the original structure. Second, sc-NSP loosens the frequency constraint and exploits the NSC generation method of positive and negative sequential patterns mining (PNSP) (a classic NSP mining method). Furthermore, a novel pruning strategy is designed to reduce the computational complexity of sc-NSP. Finally, sc-NSP obtains the support of NSC by using the most efficient bitwise-based calculation operation. Theoretical analyses show that sc-NSP performs particularly well on data sets with a large number of elements and items in sequence. Comparison and extensive experiments along with case studies on health data show that sc-NSP is 10 times more efficient than other state-of-the-art methods, and the number of NSPs obtained is 5 times greater than other methods.

The water quality index has been universally accepted as an indicator to represent the water quality status of the surface water body comprehensively. The prevalent conventional Water Quality Indices (WQIs) suffer from limitations such as 'eclipsing' and 'ambiguity'. Artificial intelligence techniques such as artificial neural networks (ANNs) have gained importance to overcome the limitations of conventional WQIs. In the present study, the Levenberg Marquardt (LM) algorithm and the Scaled Conjugate Gradient (SCG) algorithm have been compared to develop WQI based on the ANN approach (i.e ANNWQI). It is observed that the LM algorithm outperforms the SCG algorithm for prediction of ANNWQI of Indian streams, while the Bayesian Regularization algorithm has not been found suitable for the same purpose in the present study. It is also observed that both LM and SCG algorithm gives robust predictions when the hidden layer contains ten neurons. The combination of data set partitioning of training (75%), Validation (15%), and testing (10%) have been found to give the robust performance of prediction of ANNWQI for Indian streams. The predicted ANNWQI model using the LM algorithm has a very high correlation with the measured WQI values and therefore recommended to be adopted as an effective alternative, to avoid lengthy calculations involved in prevalent conventional WQI.

Clustering taxi data is commonly used to understand spatial patterns of urban mobility. In this paper, we propose a new clustering model called Origin-Destination-means (OD-means). OD-means is a hierarchical adaptive k-means algorithm based on origin-destination pairs. In the first layer of the hierarchy, the clusters are separated automatically based on the variation of the within-cluster distance of each cluster until convergence. The second layer of the hierarchy corresponds to the sub clustering process of small clusters based on the distance between the origin and destination of each cluster. The algorithm is tested on a large data set of taxi GPS data from Santiago, Chile, and compared to other clustering algorithms. In contrast to them, our proposed model is capable of detecting general and local travel patterns in the city due to its hierarchical structure.

Attribute reduction, also called feature selection, is one of the most important issues of rough set theory, which is regarded as a vital preprocessing step in pattern recognition, machine learning, and data mining. Nowadays, high-dimensional mixed and incomplete data sets are very common in real-world applications. Certainly, the selection of a promising feature subset from such data sets is a very interesting, but challenging problem. Almost all of the existing methods generated a cover on the space of objects to determine important features. However, some tolerance classes in the cover are useless for the computational process. Thus, this article introduces a new concept of stripped neighborhood covers to reduce unnecessary tolerance classes from the original cover. Based on the proposed stripped neighborhood cover, we define a new reduct in mixed and incomplete decision tables, and then design an efficient heuristic algorithm to find this reduct. For each loop in the main loop of the proposed algorithm, we use an error measure to select an optimal feature and put it into the selected feature subset. Besides, to deal more efficiently with high-dimensional data sets, we also determine redundant features after each loop and remove them from the candidate feature subset. For the purpose of verifying the performance of the proposed algorithm, we carry out experiments on data sets downloaded from public data sources to compare with existing state-of-the-art algorithms. Experimental results showed that our algorithm outperforms compared algorithms, especially in classification accuracy.

In the process industry, it is essential to establish a data-driven soft sensor to predict the key variable that is difficult to online measure directly. The accuracy performance of data-driven soft sensors relies heavily on data. Unfortunately, it is hard to acquire sufficient and informative data from the samples with limited number, which is called as the small sample problem. For handling the small sample problem, it is a good solution to generating virtual samples according to the distribution of original data. This paper proposes an enhanced method of virtual sample generation utilizing manifold features to develop soft sensors using small data. First, T-Distribution Stochastic Neighbor Embedding (t-SNE) is utilized to extract the features of input data. The main idea of generating virtual samples is to use the interpolation algorithm to obtain virtual t-SNE input features and then the random forest algorithm is utilized to get the virtual outputs using virtual t-SNE input features. Finally, virtual samples using the proposed t-SNE based virtual sample generation (t-SNE-VSG) can be achieved. For the sake of confirming the effectiveness and feasibility of the presented t-SNE-VSG, a standard data set is first used. What is more, a small data set from an actual industrial process of Purified Terephthalic Acid is used to establish a soft sensor model. The results from simulations show that the accuracy performance of the soft sensor established with small data can be effectively improved by adding the virtual samples generated by t-SNE-VSG. In addition, t-SNE-VSG achieves superior accuracy to state-of-the-art virtual sample generation methods.

This paper presents a spatial decision support tool that implements the Ordered Weighted Averaging (OWA) method. OWA is a family of multicriteria evaluation operators characterised by two sets of weights: criterion importance weights and order weights. We propose a highly interactive way of choosing, modifying, and fine-tuning the decision strategy defined by the order weights. This exploratory approach to OWA is supported by a graphical representation of the operator's behaviour in terms of decision risk and tradeoff/dispersion between criteria. Our prototype implementation is based on the CommonGIS software, and thus, Web-enabled and working with vector data. We successfully demonstrate online, exploratory support of spatial decision strategies using a data set of skiing resorts in Wallis, Switzerland.This is a post-peer-review, pre-copyedit version of an article published in Journal of Geographical Systems. The final authenticated version is available online at: http://dx.doi.org/10.1007/s101090300095

Pattern mining is a fundamental technique of data mining to discover interesting correlations in the data set. There are several variations of pattern mining, such as frequent itemset mining, sequence mining, and high utility itemset mining. High utility itemset mining is an emerging data science task, aims to extract knowledge based on a domain objective. The utility of a pattern shows its effectiveness or benefit that can be calculated based on user priority and domain-specific understanding. The sequential pattern mining (SPM) issue is much examined and expanded in various directions. Sequential pattern mining enumerates sequential patterns in a sequence data collection. Researchers have paid more attention in recent years to frequent pattern mining over uncertain transaction dataset. In recent years, mining itemsets in big data have received extensive attention based on the Apache Hadoop and Spark framework. This paper seeks to give a broad overview of the distinct approaches to pattern mining in the Big Data domain. Initially, we investigate the problem involved with pattern mining approaches and associated techniques such as Apache Hadoop, Apache Spark, parallel and distributed processing. Then we examine major developments in parallel, distributed, and scalable pattern mining, analyze them in the big data perspective and identify difficulties in designing the algorithms. In particular, we study four varieties of itemsets mining, i.e., parallel frequent itemsets mining, high utility itemset mining, sequential patterns mining and frequent itemset mining in uncertain data. This paper concludes with a discussion of open issues and opportunity. It also provides direction for further enhancement of existing approaches.

Despite several attempts to compare and examine the predictive accuracy of real estate sales and rent prices between the regression-based and neural-network (NN)-based approaches, the results are largely mixed. Prior study limitations include a small sample size and a disregard for spatial dependence, which is an essential characteristic of real estate properties. Hence, this study aims to add new empirical evidence to the literature on comparing regression-based with NN-based rent price prediction models through sophistications by (1) examining different and relatively large-scale sample sizes (n = 10(4), 10(5), 10(6)), and (2) considering the spatial dependence of either the application of nearest neighbor Gaussian processes (NNGP) or the latitude-longitude coordinate function (in the case of a deep neural network [DNN]). A case study of apartment rent prices in Japan shows that, given an increase in sample size, the out-of-sample predictive accuracies of the DNN approaches and that of NNGP are nearly equal in the order of n = 10(6). However, the DNN may have higher predictive accuracy than the NNGP for both higher- and lower-end properties whose rent prices deviate from the median.

Building footprint delineation from remote sensing imagery is a basic task in surveying and mapping and geographic information system (GIS), which benefits many engineering applications but requires an enormous amount of human delineation. In this study, we aim to find a way to replace human delineation with automated algorithms. For this purpose, we designed a novel pipeline for the automated production of building vector maps from aerial images, which consists of a bounding-box generation module, a graph convolutional network (GCN)-based polygon prediction module, and an empirical polygon regularization module. First, we introduce the bounding box generation module based on region-based object detection, which along with our overlap cropping strategy is used to generate a bounding box for each building instance. The generated bounding boxes have a horizontal rectangle version and a rotated version, which are used to initialize the next stage of our method. Second, we propose a GCN-based method, which is the core of this study, to conduct the initial accurate building polygon prediction by integrating multiresolution optimization and multilevel loss constraints. Our proposed method, to the authors' best knowledge, is the first of its kind that introduces GCN to building extraction. The final step is to apply a regularization algorithm to translate the predicted polygons into structured and highly accurate building footprints. We validated the proposed method on two large aerial building data sets, WHU data set and Inria data set, where it was shown to significantly outperform other state-of-the-art methods more than 10%. Specifically, with the ground-truth rotated bounding boxes of buildings, our method is able to automatically delineate 91% of the buildings in the WHU data set and with the predicted rotated bounding the percent reached human-level delineation.

Traditional band selection methods only analyze one data set at a time, and start searching band subsets from the zero ground state of knowledge, which can not effectively mine spectral information to guide band selection. However, for hyperspectral images obtained by the same sensor, the spectral information has similar physical meaning (radiance or reflectivity). Collaborative analysis technology can analyze multiple hyperspectral datasets to explore the inherent spectral features shared among them. In this paper, a multiple data sets collaborative analysis framework for hyperspectral band selection is proposed to realize spectral information communication, thereby guiding and promoting band selection of each data set. Different band selection tasks are established pertinently, then the evolutionary multitasking band selection method is designed to facilitate the knowledge sharing of different band selection tasks. More importantly, the interaction mechanism among different data sets is adjusted dynamically, thereby improving the cooperation ability of the collaborative analysis framework. Besides, a predominant gene reservation crossover and a deduplication mutation are designed for retaining the promising bands and avoiding the selection of repeat bands. Experiments indicate that the proposed collaborative analysis method works more efficiently than the comparison methods and successfully enhances accuracy and convergence compared to single data set analysis.

Content-weighted compression scheme for high-resolution remote-sensing (RS) images can be well modeled by Markov random field (MRF)-oriented attention. This article addresses high-resolution RS image compression by incorporating MRF into attention mechanism. To this end, we reformulate the attention mechanism with MRF-based probabilistic graph modeling implicitly and combine the target of image compression and parameter learning of MRF in a unified framework, namely high-order MRF-oriented attention (HMA) network. Specifically, HMA extends key-value query (KVQ) pairwise terms of the vanilla attention to high-order terms, by which the prior information could be expressed effectively to boost performance of high-resolution RS image compression. It is noted that several superiorities of HMA are listed. First, unlike the vanilla attention network that apt to yield coarse features, HMA is capable of output more pleasing decoding results. Second, HMA can accelerate the convergence in the training of the deep neural networks (DNNs), thus facilitating deploying it on resource-limited IOT devices. Third, HMA demonstrates its potential of processing semantic joint task. Moreover, We thoroughly evaluate our approach on standard data sets of varying resolutions, the proposed framework performs favorably against most image coding standards and DNN-based codecs on the ISPRS Vaihingen data set and the USC-SIPI data set especially at low bit rates.

The success of deep learning methods led to significant breakthroughs in 3-D point cloud processing tasks with applications in remote sensing. Existing methods utilize convolutions that have some limitations, as they assume a uniform input distribution and cannot learn long-range dependences. Recent works have shown that adding attention in conjunction with these methods improves performance. This raises a question: can attention layers completely replace convolutions? This letter proposes a fully attentional model--Point Transformer (PT) for deriving a rich point cloud representation. The model's shape classification and retrieval performance are evaluated on a large-scale urban data set--RoofN3D and a standard benchmark data set ModelNet40. Extensive experiments are conducted to test the model's robustness to unseen point corruptions for analyzing its effectiveness on real data sets. The proposed method outperforms other state-of-the-art models in the RoofN3D data set, gives competitive results in the ModelNet40 benchmark, and shows high robustness to various unseen point corruptions. Furthermore, the model is highly memory and space-efficient when compared to other methods.

Learning causal relationships from point processes is of great significance to various real-world applications, e.g., user behaviour study, fault diagnosis. Though several methods have been proposed for this problem, the existing methods rely on the stationarity assumption of the point process. Such a stationarity assumption is usually violated due to the influence of latent confounders of the point processes. Based on the study of various real-world point processes, we find that a non-stationary Hawkes process is usually a mixture of several non-overlap and stationary processes. Thus, we propose an adaptive pattern based method for the non-stationary Hawkes Process (named GC-nsHP). In the proposed method, the following two steps are iteratively employed to adaptively partition the non-stationary processes and learn the causal structure for the partitioned sub-processes: (1) we use a dynamic-programming-based algorithm to partition the non-stationary long process into several stationary sub-processes; (2) we use an expectation–maximization-based algorithm (EM) to learn the Granger Causality of each pattern. Experiments on both synthetic and real-world datasets not only show the effectiveness of the proposed method on the non-stationary point process, but also discover some interesting results on the IPTV data set.

The Multilayer Perceptron produces the high level of accuracy i.e., 72.7% accuracy level. The Logistic approach produces the 81.45% of precision value. This is the highest precision value compare with other model. Rest of the algorithms namely Gaussian Processes, Linear Regression, Multilayer Perceptron and SGD have produce the precision values from 71 to 81%. The Multilayer Perceptron approach takes low time consumption like 0.09 s to build the model, the Next SGD approach takes 0.12 s to build the model, the Next Logistic approach takes the 0.18 s, the Next Gaussian Processes approach takes the 0.2 s and finally Linear Regression approach takes more time around 0.21 s to build the model. The Linear Regression, Multilayer Perceptron have produces the recall value most nearly with Gaussian Processes. Rest of the approaches like SGD results vast difference with our optimal results and Logistic has very low and vast difference with our optimal approaches. AI strategies are used in an ever-increasing number of true applications including limitations on computational spending plans, the precise incorporation of such imperatives into the cycle of model choice, and model improvement is needed to an expanding degree.

A large number of clustering methods exist including deterministic, probabilistic and fuzzy clustering. All these methods are devoted to handling different types of uncertainty. No studies have been encountered on clustering taking into account a confluence of probabilistic and fuzzy information. In the existing studies, the reliability of extracted knowledge is one of the important issues to be investigated. The concept of Z-number arises as a formal construct that expresses reliability of information under bimodal distribution. In this paper, we propose an approach to construction of Z-number-valued clusters of a data set for evaluation of reliability of extracted data-driven knowledge. Real-world applications are given that confirm the usefulness of the proposed method.

Uncertainty measure is an important tool for data analysis. In practical applications, the collected data are subject to different probability distributions. This requires that the uncertainty measure has generalization performance. Fuzzy rough set (FRS) theory is a popular mathematical tool for uncertainty measure, but the theory does not work well for some data distributions. For example, when the class density difference of the data set is large, FRS theory cannot effectively evaluate the classification uncertainty of samples. In this study, we combine the relative measure with the lower approximation of FRSs to propose a relative uncertainty measure which can address the above-mentioned problem. Furthermore, a fuzzy rough feature selection algorithm is designed, and it is mainly used to test the effectiveness and efficiency of the proposed measure. Experimental results demonstrate that the proposed feature selection algorithm has good performance. It indirectly proves that the relative uncertainty measure is effective and efficient in classification tasks.

The development of body sensor networks (BSNs) with rich multimodal signals has enabled highly accurate fine-grained action detection, which is the cornerstone of many human–computer interaction applications. However, in the case of consecutive fine-grained actions, most existing wearable sensor-based detection methods are constrained by sliding windows because of their limited temporal receptive fields, and existing sequence-to-sequence detection methods cannot effectively leverage the potential of multimodal information of wearable sensors. Herein, to give multimodal signals full play in fine-grained action detection, we propose a novel temporal convolutional network by designing a channel attention-based multistream structure. We apply it to a promising application for correct and incorrect patient transfer nursing action detection. A data set is collected from a BSN on a patient when nurses perform patient transfer. Extensive experiments on our data set and public data set (C-MHAD) demonstrate that the proposed method is superior to the state-of-the-art methods, because it can strengthen the utilization of prediction features from the more convincing modal stream at each time frame. The source code is available at https://github.com/zzh-tech/Continuous-Action-Detection .

Traffic data collected from sensor networks often exhibit strong spatial correlations and recurrent temporal patterns. Learning these patterns and diagnosing anomalies in such spatiotemporal traffic data is critical to improving transportation systems and services. This paper proposes a dynamic framework to model spatiotemporal traffic data, with a particular application on diagnosing anomalies. Within the framework, we focus on characterizing the variation in system dynamics with a time-varying vector autoregressive model. We impose a low-rank tensor structure to model the collection of time-varying system matrices. As the temporal factor matrix captures the principal patterns/signatures across all time-varying system matrices, it is a useful tool to diagnose abnormal generative mechanisms and unexpected temporal patterns. We demonstrate the proposed tensor learning framework's effectiveness by experimenting with a synthetic data set and real-world spatiotemporal traffic speed data set. The results show the superiority of the proposed model in uncovering anomalous traffic network dynamics.

Background A complete dataset is essential for biomedical implementation. Due to the limitation of objective or subjective factors, missing data often occurs, which exerts uncertainty in the subsequent data processing. Commonly used methods of interpolation are interpolating substitute values that keep minimum error. Some applications of statistics are usually used for handling this problem. Methods We are trying to find a higher performance interpolation method compared with the usual statistic methods, by using artificial intelligence which is in full swing today. The prediction and classification of backpropagation neural network are used in this paper, describes a missing data interpolation method to propose the interpolation model that mines association rules in the data. In the experiment, depending on a multi-layer network structure, the model is trained and tested by sample data, constantly revises network weights and thresholds. The error function decreases along the negative gradient direction and approaches the expected real output. The model is validated on the breast cancer dataset, and we select real samples from the data set for validation, moreover, add four traditional methods as a control group. Results The proposed method has great performance improvement in the interpolation of missing data. Experimental results show that the interpolation accuracy of our proposed method (84%) is higher than four traditional methods (1.33%, 74.67%, 73.33%, 77.33%) as mentioned in this paper, BPNN stays low in MSE evaluation. Finally, we analyze the performance of various methods in processing missing data. Conclusions The study in this paper has estimated missing data with high accuracy as much as possible to reduce the negative impact in the diagnosis of real life. At the same time, it can also assist in missing data processing in the biomedical field.

In the context of big data analysis, the divide-and-conquer methodology refers to a multiple-step process: first splitting a data set into several smaller ones; then analyzing each set separately; finally combining results from each analysis together. This approach is effective in handling large data sets that are unsuitable to be analyzed entirely by a single computer due to limits either from memory storage or computational time. The combined results will provide a statistical inference which is similar to the one from analyzing the entire data set. This article reviews some recently developments of divide-and-conquer methods in a variety of settings, including combining based on parametric, semiparametric and nonparametric models, online sequential updating methods, among others. Theoretical development on the efficiency of the divide-and-conquer methods is also discussed.

Basic oxygen furnace (BOF) steelmaking end-point control using soft measurement models has essential value for economy and environment. However, the high-dimensional and redundant data of the BOF collected by the sensors will hinder the performance of models. The traditional feature selection results based on meta-heuristic algorithms cannot meet the stability of actual industrial applications. In order to eliminate the negative impact of feature selection application in the BOF steelmaking, an improved grey wolf optimizer (IGWO) for feature selection was proposed, and it was applied to the BOF data set. Firstly, the proposed algorithm preset the size of the feature subset based on the new encoding scheme, rather than the traditional uncertain number strategy. Then, opposition-based learning was used to initialize the grey wolf population so that the initial population was closer to the potential optimal solution. In addition, a novel population update method retained the features closely related to the best three grey wolves and probabilistically updated irrelevant features through measurement or random methods. These methods were used to search feature subsets to maximize search capability and stability of algorithm on BOF steelmaking data. Finally, the proposed algorithm was compared with other feature selection algorithms on the BOF data sets. The results show that the proposed IGWO can stably select the feature subsets that are conductive to the end-point regression accuracy control of BOF temperature and carbon content, which can improve the performance of the BOF steelmaking.

Data-based methods are capable to monitor machine components. Approaches for semi-supervised anomaly detection are trained using sensor data that describe the normal state of machine components. Thus, such approaches are interesting for industrial practice, since sensor data do not have to be labeled in a time-consuming and costly way. In this work, an ensemble approach for semi-supervised anomaly detection is used to detect anomalies. It is shown that the ensemble approach is suitable for condition monitoring of ball screws. For the evaluation of the approach, a data set of a regular test cycle of a ball screw from automotive industry is used.

Aiming at the shortcomings of a single machine learning model with low model prediction accuracy and insufficient generalization ability in house price index prediction, a whale algorithm optimized support vector regression model based on bagging ensemble learning method is proposed. Firstly, gray correlation analysis is used to obtain the main influencing factors of house prices, and the segmentation forecasting method is used to divide the data set and forecast the house prices in the coming year using the data of the past ten years. Secondly, the whale optimization algorithm is used to find the optimal parameters of the penalty factor and kernel function in the SVR model, and then, the WOA-SVR model is established. Finally, in order to further improve the model generalization capability, a bagging integration strategy is used to further integrate and optimize the WOA-SVR model. The experiments are conducted to forecast the house price indices of four regions, Beijing, Shanghai, Tianjin, and Chongqing, respectively, and the results show that the prediction accuracy of the proposed integrated model is better than the comparison model in all cases.

Phylogenetic placement of query samples on an existing phylogeny is increasingly used in molecular ecology, including sample identification and microbiome environmental sampling. As the size of available reference trees used in these analyses continues to grow, there is a growing need for methods that place sequences on ultra-large trees with high accuracy. Distance-based placement methods have recently emerged as a path to provide such scalability while allowing flexibility to analyze both assembled and unassembled environmental samples. In this paper, we introduce a distance-based phylogenetic placement method, APPLES-2, that is more accurate and scalable than existing distance-based methods and even some of the leading maximum likelihood methods. This scalability is owed to a divide-and-conquer technique that limits distance calculation and phylogenetic placement to parts of the tree most relevant to each query. The increased scalability and accuracy enables us to study the effectiveness of APPLES-2 for placing microbial genomes on a data set of 10,575 microbial species using subsets of 381 marker genes. APPLES-2 has very high accuracy in this setting, placing 97% of query genomes within three branches of the optimal position in the species tree using 50 marker genes. Our proof of concept results show that APPLES-2 can quickly place metagenomic scaffolds on ultra-large backbone trees with high accuracy as long as a scaffold includes tens of marker genes. These results pave the path for a more scalable and widespread use of distance-based placement in various areas of molecular ecology.

IoT sensors are becoming increasingly important supplement to traditional monitoring systems, particularly for in-situ based monitoring. Data collected using IoT sensors are often plagued with missing values occurring as a result of sensor faults, network failures, drifts and other operational issues. Missing data can have substantial impact on in-field sensor calibration methods. The goal of this research is to achieve effective calibration of sensors in the context of such missing data. To this end, two objectives are presented in this paper. 1) Identify and examine effective imputation strategy for missing data in IoT sensors. 2) Determine sensor calibration performance using calibration techniques on data set with imputed values. Specifically, this paper examines the performance of Variational Autoencoder (VAE), Neural Network with Random Weights (NNRW), Multiple Imputation by Chain Equations (MICE), Random Forest-based Imputation (missForest) and K-Nearest Neighbour (KNN) for imputation of missing values on IoT sensors. Furthermore, the performance of sensor calibration via different supervised algorithms trained on the imputed dataset were evaluated. The analysis showed VAE technique to outperform the other methods in imputing the missing values at different proportions of missingness on two real-world datasets. Experimental results also showed improved calibration performance with imputed dataset.

Fire detection and alarm system is fully concerned for safety. And convolutional neural network (CNN) has been introduced into fire/smoke detection based on video/image understanding. However, the samples of the existed public fire/smoke data sets are not enough to train very deep CNN. And the generalization abilities of the existed methods are limited. Therefore, a fire monitoring and alarm system (FMAS) based on channel-wise pruned YOLOv3 is proposed, and a big fire and smoke image data set has been collected in this paper. YOLOv3 with Darknet-53 is deep and its generalization ability has been demonstrated in general objects detection. But it has massive parameters, which may hinder its applications in fire monitoring systems with restricted computation resources. Thus, channel-wise pruning technology is introduced to reduce the number of parameters while bringing a slight drop of accuracy. Moreover, OHEM technology is proposed to improve the detection accuracy further. Multiple comparison experiments on the homemade data set and the public data sets have demonstrated that the proposed channel-wise pruned YOLOv3 with OHEM can achieve satisfactory accuracy with low calculation capacity after squeezing parameters.

For the problems of low accuracy and low efficiency of most load forecasting methods, a load forecasting method based on improved deep learning in cloud computing environment is proposed. Firstly, the preprocessed data set is divided into several data partitions with relatively balanced data volume through spatial grid, so as to better detect abnormal data. Then, the density peak clustering algorithm based on spark is used to detect abnormal data in each partition, and the local clusters and abnormal points are merged. The parallel processing of data is realized by using spark cluster computing platform. Finally, the deep belief network is used for load classification, and the classification results are input into the empirical mode decomposition-gating recurrent unit network model, and the load prediction results are obtained through learning. Based on the load data of a power grid, the experimental results demonstrate that the mean prediction error of the proposed method is basically controlled within 3% in the short term and 0.023 MW, 19.75%, and 2.76% in the long term, which are better than other comparison methods, and the parallel performance is good, which has a certain feasibility.

Traditional machine learning methods assume that training and testing data must be from the same machine running condition (MRC) and drawn from the same distribution. However, in several real-time industrial applications, this assumption does not hold. The traditional methods work satisfactorily in steady-state conditions but fail in time-varying conditions. In order to utilize time-varying data in variable MRCs, this article proposes a novel low-level knowledge transfer framework using a deep neural network (DNN) model for condition monitoring of machines in variable running conditions. The low-level features have been extracted in time, frequency, and time–frequency domains. These features are extracted from the source data to train the DNN. The trained DNN-based parameters are then transferred to another DNN, which is modified according to the low-level features extracted from the target data. The proposed approach is validated through three case studies on: 1) the air compressor acoustic data set; 2) the Case Western Reserve University bearing data set; and 3) the intelligent maintenance system bearing data set. The prediction accuracy obtained for the above case studies is as high as 100%, 93.07%, and 100%, respectively, with fivefold cross-validation. These real-time results show considerable improvement in the prediction performance using the proposed approach. Note to Practitioners —Condition-based monitoring schemes are widely applicable to rotating machines in various industries since they operate in tough working situations, and consequently, unpredicted failures occur. These unpredicted failures may cause perilous accidents in the industries. CBM systems prevent such failures, which results in the reduction of equipment damage and, hence, increases machinery lifetime. Modern industries are so complex and generating huge data, and these data can be collected using sensors, but placing a large number of sensors is difficult and expensive for different but similar kinds of faults in industries. This also increases the cost due to additional sensors and circuits. In this article, the authors have proposed a novel low-level knowledge transfer framework using the deep neural network (DNN)-based method for condition monitoring of machines in variable running conditions. Low-level features have been extracted to reduce the computations of DNN drastically with improved performance. This article also considered additional faults in the target domain, which is more practical in real-time applications. The proposed scheme has been validated with three case studies on acoustic and vibration signatures.

Crop cultivation prediction is an integral part of agriculture and is primarily based on factors such as soil, environmental features like rainfall and temperature, and the quantum of fertilizer used, particularly nitrogen and phosphorus. These factors, however, vary from region to region: consequently, farmers are unable to cultivate similar crops in every region. This is where machine learning (ML) techniques step in to help find the most suitable crops for a particular region, thus assisting farmers a great deal in crop prediction. The feature selection (FS) facet of ML is a major component in the selection of key features for a particular region and keeps the crop prediction process constantly upgraded. This work proposes a novel FS approach called modified recursive feature elimination (MRFE) to select appropriate features from a data set for crop prediction. The proposed MRFE technique selects and ranks salient features using a ranking method. The experimental results show that the MRFE method selects the most accurate features, while the bagging technique helps accurately predict a suitable crop. The performance of proposed MRFE technique is evaluated by various metrics such as accuracy (ACC), precision, recall, specificity, F1 score, area under the curve, mean absolute error, and log loss. From the performance analysis, it is justified that the MRFE technique performs well with 95% ACC than other FS methods.

The partitioning-based k-means clustering is one of the most important clustering algorithms. However, in big data environment, it faces the problems of random selection of initial cluster centers randomly, expensive communication overhead among MapReduce nodes and data skewing in data partitions, and others. To solve these problems, this paper proposes a parallel clustering algorithm based on grid density and local sensitive hash function (MR-PGDLSH) which takes into account the advantages of MapReduce and LSH (locality sensitive hash function). In the MR-PGDLSH, firstly the GDS (grid density strategy) is designed to obtain the relatively reasonable initial cluster centers. Then, a DP-LSH (data partition based on locality sensitive hash function) is proposed to divide the data set into multiple segments. The relevant data objects are mapped to the same sub-data set. The similarity function is designed to generate clusters, thereby reducing frequent communication overhead between nodes. Next, the AGS (adaptive grouping strategy) is applied to distribute the amount of data on each node evenly, which solves the problem of data skew on the node. Finally, the MR-PGDLSH is applied to mine the cluster centers in parallel, which obtains the final clustering results. Both theoretical analysis and experimental results have shown that the MR-PGDLSH is superior to the existing clustering algorithms.

The number of publicly available microbiome samples is continually growing. As data set size increases, bottlenecks arise in standard analytical pipelines. Faith's phylogenetic diversity (Faith's PD) is a highly utilized phylogenetic alpha diversity metric that has thus far failed to effectively scale to trees with millions of vertices. Stacked Faith's phylogenetic diversity (SFPhD) enables calculation of this widely adopted diversity metric at a much larger scale by implementing a computationally efficient algorithm. The algorithm reduces the amount of computational resources required, resulting in more accessible software with a reduced carbon footprint, as compared to previous approaches. The new algorithm produces identical results to the previous method. We further demonstrate that the phylogenetic aspect of Faith's PD provides increased power in detecting diversity differences between younger and older populations in the FINRISK study's metagenomic data.

Uncertainty modeling of renewable energy sources, load demand, electricity price, etc. create a high volume of data in smart grids. Accordingly, in this article, a precise forecasting method based on a deep learning concept with microclustering (MC) task is presented. The MC method is structured based on hybrid unsupervised and supervised clustering tasks by $K$ -means and Gaussian support vector machine, respectively. In the proposed method, the input data sequence is clustered by the MC task, and then the forecasting process is employed. By applying the MC, input data in each hour are categorized into different groups, and a distinctive forecasting unit is allocated to each one. In this way, more clusters and forecasting networks are earmarked for the hours with higher fluctuation rates. The bi-directional long short-term memory (B-LSTM), which is one of the newest recurrent artificial neural networks, is proposed as the forecasting unit. The B-LSTM has bidirectional memory—feedforward and feedback loops—that helps us to investigate both previous and future hidden layers data. The optimal number of clusters in each hour is determined based on the Davies–Bouldin index. To evaluate the performance of the proposed method, in this study, three forecasting tasks including the wind speed, load demand, and electricity price are studied in different periods using the Ontario province, Canada, data set. The results are compared with other benchmarking methods to verify the robustness and effectiveness of the proposed method. In fact, the proposed method, which is equipped with the MC technique and B-LSTM networks, significantly promotes the forecasting results, especially in spike points.

In this paper, a hybrid model with both clustering mechanism and regression mechanism is built with aid of an allocation of information granularity . Some numeric prototypes constructed by clustering methods are expanded into granular prototypes. The regression module accepts granular prototypes and outcomes information granules. Some advantages of this idea can be concluded: 1). The constructed granular prototypes are best able to represent original data with comparable smaller quantity and high quality. 2). The granular output could be used to predict a new sample’s real number output by giving a rational range. 3). It can be used to solve clustering and regression problems simultaneously. Two different granularity allocation strategies are proposed and experimented while constructing granular prototypes: non-uniformly allocation of information granularity to each cluster and non-uniformly allocation to each feature. A comprehensive objective function is defined considering specificity and generality of the output. The allocation of granularity itself is in fact a multiple-parameters optimization problem which invokes the usage of an evolutionary method. Two popular methods (GA and PSO) are tried and compared with a real data set’s experiment. Three data sets are collected from UCI website to testify the effectiveness of our approach in the experimental part.

The aim of this work is to analyse the performance of the new proposed hybrid parameterisation model in handling problematic data. Three types of problematic data will be highlighted in this paper: i) big data set, ii) uncertain and inconsistent data set and iii) imbalanced data set. The proposed hybrid model is an integration of three main phases which consist of the data decomposition, parameter reduction and parameter selection phases. Three main methods, which are soft set and rough set theories, were implemented to reduce and to select the optimised parameter set, while a neural network was used to classify the optimised data set. This proposed model can process a data set that might contain uncertain, inconsistent and imbalanced data. Therefore, one additional phase, data decomposition, was introduced and executed after the pre-processing task was completed in order to manage the big data issue. Imbalanced data sets were used to evaluate the capability of the proposed hybrid model in handling problematic data. The experimental results demonstrate that the proposed hybrid model has the potential to be implemented with any type of data set in a classification task, especially with complex data sets.

Background Missing data is a common issue in different fields, such as electronics, image processing, medical records and genomics. They can limit or even bias the posterior analysis. The data collection process can lead to different distribution, frequency, and structure of missing data points. They can be classified into four categories: Structurally Missing Data (SMD), Missing Completely At Random (MCAR), Missing At Random (MAR) and Missing Not At Random (MNAR). For the three later, and in the context of genomic data (especially non-coding data), we will discuss six imputation approaches using 31,245 variants collected from ClinVar and annotated with 13 genome-wide features. Results Random Forest and kNN algorithms showed the best performance in the evaluated dataset. Additionally, some features show robust imputation regardless of the algorithm (e.g. conservation scores phyloP7 and phyloP20), while other features show poor imputation across algorithms (e.g. PhasCons). We also developed an R package that helps to test which imputation method is the best for a particular data set. Conclusions We found that Random Forest and kNN are the best imputation method for genomics data, including non-coding variants. Since Random Forest is computationally more challenging, kNN remains a more realistic approach. Future work on variant prioritization thru genomic screening tests could largely profit from this methodology.

In December 2019, SARS-CoV-2 caused the coronavirus disease (COVID-19) to spread to all countries, infecting thousands of people and causing death. COVID-19 causes mild illness in most cases, although it can make some people seriously ill. Therefore, vaccines are in various phases of clinical progress, and some of them have been approved for national use. The current state of affairs reveals that there is a critical need for a quick and timely solution to the need for a Covid-19 vaccine. Non-clinical methods such as data mining and machine learning techniques can help to do this. This study will focus on US COVID-19 Vaccination Advances using Machine learning classification algorithms and Using Geospatial analysis to visualize the results. The paper's findings indicate which algorithm is better for a given data set. Naive Bayes algorithm is used to run tests on real world data, and is used to analyze and draw conclusions. Period of Accuracy and performance, and it was found that Naive Bayes is very superior in terms of time and accuracy.

Accurate prediction of the traffic state can help to address the issue of traffic congestion, providing guiding advices for people’s travel and traffic regulation. In this paper, we propose a novel short-term traffic flow prediction approach based on empirical mode decomposition and combination model fusion. First, we explore the amplitude-frequency characteristics of short-term traffic flow series, and use empirical mode decomposition to decompose traffic flow to several components with different frequency. Second, based on the results of self-similarity analysis of each component, improved extreme learning machine, seasonal auto regressive integrated moving average and auto regressive moving average are selected to predict different components. Meanwhile, an improved fruit fly optimization algorithm is proposed to optimize the weight coefficient of the combination model. Third, the prediction results of each prediction model are multiplied by their respective weight coefficient to get the final prediction results. We evaluate our prediction approach by doing thorough experiment on a real traffic data set. Moreover, experimental results show that the proposed approach has superior performance than state-of-the-art prediction methods or models in short-term traffic flow prediction.

In model-based clustering, the Galaxy data set is often used as a benchmark data set to study the performance of different modeling approaches. Aitkin (Stat Model 1:287–304) compares maximum likelihood and Bayesian analyses of the Galaxy data set and expresses reservations about the Bayesian approach due to the fact that the prior assumptions imposed remain rather obscure while playing a major role in the results obtained and conclusions drawn. The aim of the paper is to address Aitkin’s concerns about the Bayesian approach by shedding light on how the specified priors influence the number of estimated clusters. We perform a sensitivity analysis of different prior specifications for the mixtures of finite mixture model, i.e., the mixture model where a prior on the number of components is included. We use an extensive set of different prior specifications in a full factorial design and assess their impact on the estimated number of clusters for the Galaxy data set. Results highlight the interaction effects of the prior specifications and provide insights into which prior specifications are recommended to obtain a sparse clustering solution. A simulation study with artificial data provides further empirical evidence to support the recommendations. A clear understanding of the impact of the prior specifications removes restraints preventing the use of Bayesian methods due to the complexity of selecting suitable priors. Also, the regularizing properties of the priors may be intentionally exploited to obtain a suitable clustering solution meeting prior expectations and needs of the application.

The increasing common use of incidental unrectified satellite images have many applications for mapping of earth for coastal and ocean applications. Hazard assessment and natural resource management can also be done via this process. Remote sensing is being used extensively due to the increase in the number of satellites in space. It is also the future of optimization of GPS systems and the internet. To demonstrate the semantic segmentation process, this study presents proposed solutions along with their evaluation metrics adapted from fully connected neural networks such as UNet and PSPNet. UNet architecture based deep learning model has outperformed PSPNet based architecture with overall Mean-IOU score of 0.51 on the test set in the semantic segmentation. The overall accuracy of the model can further be improved by providing homogeneous features to train the model, balance classes and by incorporating more data set for semantic segmentation. The developed model can be useful to the authorities for smart city planning and landuse mapping.

This article presents the results of connecting an educational data mining techniques to the academic performance of students. Three classification models (Decision Tree, Random Forest and Deep Learning) have been developed to analyze data sets and predict the performance of students. The projected submission of the three classificatory was calculated and matched. The academic history and data of the students from the Office of the Registrar were used to train the models. Our analysis aims to evaluate the results of students using various variables such as the student's grade. Data from (221) students with (9) different attributes were used. The results of this study are very important, provide a better understanding of student success assessments and stress the importance of data mining in education. The main purpose of this study is to show the student successful forecast using data mining techniques to improve academic programs. The results of this research indicate that the Decision Tree classifier overtakes two other classifiers by achieving a total prediction accuracy of 97%.

The bubble point pressure (Pb) is a crucial pressure–volume–temperature (PVT) property and a primary input needed for performing many petroleum engineering calculations, such as reservoir simulation. The industrial practice of determining Pb is by direct measurement from PVT tests or prediction using empirical correlations. The main problems encountered with the published empirical correlations are their lack of accuracy and the noncomprehensive data set used to develop the model. In addition, most of the published correlations have not proven the relationships between the inputs and outputs as part of the validation process (i.e., no trend analysis was conducted). Nowadays, deep learning techniques such as long short-term memory (LSTM) networks have begun to replace the empirical correlations as they generate high accuracy. This study, therefore, presents a robust LSTM-based model for predicting Pb using a global data set of 760 collected data points from different fields worldwide to build the model. The developed model was then validated by applying trend analysis to ensure that the model follows the correct relationships between the inputs and outputs and performing statistical analysis after comparing the most published correlations. The robustness and accuracy of the model have been verified by performing various statistical analyses and using additional data that was not part of the data set used to develop the model. The trend analysis results have proven that the proposed LSTM-based model follows the correct relationships, indicating the model’s reliability. Furthermore, the statistical analysis results have shown that the lowest average absolute percent relative error (AAPRE) is 8.422% and the highest correlation coefficient is 0.99. These values are much better than those given by the most accurate models in the literature.