Search

Detecting fraud accurately in credit cards is critical as this financial sector incurs significant losses for cardholders. Nonetheless, most studies adopted standard machine learning and few incremental learning, which are inadequate for addressing credit card challenges, such as rapid data arrival, unlimited data, data sensitivity, and performance decline over time. For this purpose, we propose a chunk-based incremental feature learning approach that optimises the fraud model topology for each new chunk and keeps track of one chunk each time. The model consists of several connected sub-models, where a new sub-model is optimally created for each new chunk. To avoid the network growing indefinitely, we limit the number of sub-models. To this end, we retain the most relevant sub-models to the current chunk's data distribution and re-combine them to create the optimal model. We evaluate our approach using two credit card datasets: the first of medium scale contains 2-day payments in 2013, and the second of considerable scale possesses 6-month payments in 2019. We split these datasets into multiple chunks to learn and test incrementally. We compare our approach with static learning methods trained with different scenarios. Moreover, we vary the number of historical sub-models to check their impact on the predictive performance. [ABSTRACT FROM AUTHOR]

Every second, a huge volume of multi-dimensional data is generated in fields such as Social Networking, Industrial Internet of Things, Stock market and E-commerce applications. Knowledge and pattern extraction are a challenging task in the evolving nature of data stream. Major issues are (i) ‘concept drift’ occurs as a result of pattern changes in the data distribution and (ii) ‘concept evolution’ occurs when a new class evolves in the data stream. These issues degrade the performance of learning models. In this paper, we focus on detection of concept evolution and enhance the performance of classifiers. For this, we propose a new model to identify novel classes, namely, Detection of Novel Classes (DNC). The proposed method adopts long short term memory to continuously observe the streaming data in order to detect emerging classes. The continuous monitoring allows the model to distinguish between existing classes and the novel classes which save time and memory. Also, the proposed method is demonstrated for identifying more than one novel class. The experiments are performed over seven different datasets. The results confirm the efficiency is increased ranging from 6% to 34% by the proposed method in identifying new concepts in the evolving data stream than the existing methods available in the literature. [ABSTRACT FROM AUTHOR]

Concept drift in data stream processing remains an intriguing challenge and states a popular research topic. Methods that actively process data streams usually employ drift detectors, whose performance is often based on monitoring the variability of different stream properties. This publication provides an overview and analysis of metafeatures variability describing data streams with concept drifts. Five experiments conducted on synthetic, semi-synthetic, and real-world data streams examine the ability of over 160 metafeatures from 9 categories to recognize concepts in non-stationary data streams. The work reveals the distinctions in the considered sources of streams and specifies 17 metafeatures with a high ability of concept identification. [ABSTRACT FROM AUTHOR]

The distribution of data often evolves over time, necessitating classifiers to adjust in order to maintain optimal classification accuracy. This phenomenon, termed "concept drift", poses a significant challenge. Detectors specifically designed for identifying concept drift are typically integrated to bolster classifier performance. In this study, we introduce two innovative methodologies designed to address the prevalent issues of high miss detection, excessive false alarms, and prolonged detection latencies encountered in many contemporary concept drift detection algorithms. The first approach, termed the Hybrid Weighting-based Concept Drift Detection Method (HW_DDM), incorporates both linear and exponential weighting for long and short windows, respectively, within a composite window model. Subsequently, concept drift is detected by calculating the weighted mean value within the window, leveraging the Hoeffding and McDirmid inequality thresholds. The second strategy, named the Dynamic Weighting-based Hoeffding Drift Detection Method (DW_HDDM), employs a mechanism that dynamically adjusts the Hoeffding threshold and dynamically weights classification prediction outcomes, thereby catering to the drift and augmenting detection efficacy. Comparative evaluations using both synthetic and real-world datasets against leading-edge algorithms are presented. The empirical results underscore that HW_DDM exhibits the lowest false detection rate with negligible miss detections on synthetic datasets. In contrast, DW_HDDM shines with minimal detection delay, reduced miss detection rates, and a diminished false detection rate, demonstrating superior classification accuracy on real-world datasets when pitted against benchmark algorithms. [ABSTRACT FROM AUTHOR]

The development in the fields of clean energy, particularly wind and photovoltaic power, generates a large amount of data streams, and how to mine valuable information from these data to improve the efficiency of power generation has become a hot spot of current research. Traditional classification algorithms cannot cope with dynamically changing data streams, so data stream classification techniques are particularly important. The current data stream classification techniques mainly include decision trees, neural networks, Bayesian networks, and other methods, which have been applied to wind power and photovoltaic power data processing in existing research. However, the data drift problem is gradually highlighted due to the dynamic change in data, which significantly impacts the performance of classification algorithms. This paper reviews the latest research on data stream classification technology in wind power and photovoltaic applications. It provides a detailed introduction to the data drift problem in machine learning, which significantly affects algorithm performance. The discussion covers covariate drift, prior probability drift, and concept drift, analyzing their potential impact on the practical deployment of data stream classification methods in wind and photovoltaic power sectors. Finally, by analyzing examples for addressing data drift in energy-system data stream classification, the article highlights the future prospects of data drift research in this field and suggests areas for improvement. Combined with the systematic knowledge of data stream classification techniques and data drift handling presented, it offers valuable insights for future research. [ABSTRACT FROM AUTHOR]

In this paper, a novel autonomous centreless algorithm is proposed for data partitioning. The proposed algorithm firstly constructs the nearest neighbour affinity graph and identifies the local peaks of data density to build micro-clusters. Unlike the vast majority of partitional clustering algorithms, the proposed algorithm does not rely on singleton prototypes, namely, centres or medoids of the micro-clusters to partition the data space. Instead, these micro-clusters are directly utilised to attract nearby data samples to form shape-free Voronoi tessellations, hence, being centreless and robust to noisy data. A fusion scheme is further implemented to fuse these data clouds with higher intra-cluster similarity together to attain a more compact partitioning of data. The proposed algorithm is able to perform data partitioning on a chunk-wise basis and is highly computationally efficient with the default distance measure. Therefore, it is suitable for both static data partitioning in offline scenarios and streaming data partitioning in online scenarios. Numerical examples on a variety of benchmark datasets demonstrate the efficacy of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Learning from changing data streams is one of the important tasks of data mining. The phenomenon of the underlying distribution of data streams changing over time is called concept drift. In classification decision-making, the occurrence of concept drift will greatly affect the classification efficiency of the original classifier, that is, the old decision-making model is not suitable for the new data environment. Therefore, dealing with concept drift from changing data streams is crucial to guarantee classifier performance. Currently, most concept drift detection methods apply the same detection strategy to different data streams, with little attention to the uniqueness of each data stream. This limits the adaptability of drift detectors to different environments. In our research, we designed a unique solution to address this issue. First, we proposed a variance estimation strategy and a variance feedback strategy to characterize the data stream's characteristics through variance. Based on this variance, we developed personalized drift detection schemes for different data streams, thereby enhancing the adaptability of drift detection in various environments. We conducted experiments on data streams with various types of drifts. The experimental results show that our algorithm achieves the best average ranking for accuracy on the synthetic dataset, with an overall ranking 1.12 to 1.5 higher than the next-best algorithm. In comparison with algorithms using the same tests, our method improves the ranking by 3 to 3.5 for the Hoeffding test and by 1.12 to 2.25 for the McDiarmid test. In addition, they achieve a good balance between detection delay and false positive rates. Finally, our algorithm ranks higher than existing drift detection methods across the four key metrics of accuracy, CPU time, false positives, and detection delay, meeting our expectations. [ABSTRACT FROM AUTHOR]

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

The traditional high utility itemsets mining algorithms cannot discover the relationships between categories at different levels of abstraction. Therefore, cross-level high utility itemsets mining algorithms have been proposed. To address the problem that the current cross-level high utility itemsets mining algorithms can only handle static data and cannot control the range of mining levels, a dynamic classification list structure DTUL is proposed to store and maintain the utility and classification information of the itemset in the window. Based on this structure, this paper proposes the first constrained cross-level high utility itemset mining algorithm based on sliding window, including the bottom-up mining algorithm CCLHM_DTU algorithm and the top-down mining algorithm CCLHM_UTD algorithm. Extensive experiments are conducted on databases containing categorical information. The experimental results show that the proposed algorithm can effectively deal with the data stream and flexibly constrain the hierarchical range of itemsets. [ABSTRACT FROM AUTHOR]

Anomaly detection is a challenging task in surveillance videos. Nowadays, deep learning-based approaches have been developed for this issue. Although they achieve good performance, the training procedure is time-consuming and expensive due to including many layers. Besides, they have a dependence on the training set, and the structure of the deep networks is not suitable for dynamic events. They require retraining to detect different normal or abnormal events. They cannot adapt to the new event dynamically and instantly without retraining. This paper aims to bring a novel perspective to anomaly detection in surveillance video by tackling the task in a data stream manner to overcome these disadvantages. A novel and simple data stream-based ensemble approach for video anomaly detection is presented in this paper. Initially, fixed-sized temporal segments are created using current frames during streaming. A multiple instance learning-based preprocessing method is applied to the current segment, and a 1-D flow vector is obtained. A fixed-sized chunk is generated by pooling the flow vectors for learning. Afterwards, a streaming data learning algorithm called Unsupervised Feature Representative Online Sequential-Extreme Learning Machines (UFROS-ELM) is applied to the current chunk. UFROS-ELM makes the initial prediction about the current vectors using a concept drift detection mechanism and ELM-based autoencoder. Finally, multiple UFROS-ELM based ensemble learning is employed for the final decision using the majority voting. The results are achieved on the well-known surveillance data sets and compared with state-of-the-art deep learning-based video anomaly detection algorithms. The promising results support further research in this area. [ABSTRACT FROM AUTHOR]

In nonstationary environments, high-dimensional data streams have been generated unceasingly where the underlying distribution of the training and target data may change over time. These drifts are labeled as concept drift in the literature. Learning from evolving data streams demands adaptive or evolving approaches to handle concept drifts, which is a brand-new research affair. In this effort, a wide-ranging comparative analysis of concept drift is represented to highlight state-of-the-art approaches, embracing the last four decades, namely from 1980 to 2020. Considering the scope and discipline; the core collection of the Web of Science database is regarded as the basis of this study, and 1,564 publications related to concept drift are retrieved. As a result of the classification and feature analysis of valid literature data, the bibliometric indicators are revealed at the levels of countries/regions, institutions, and authors. The overall analyses, respecting the publications, citations, and cooperation of networks, are unveiled not only the highly authoritative publications but also the most prolific institutions, influential authors, dynamic networks, etc. Furthermore, deep analyses including text mining such as; the burst detection analysis, co-occurrence analysis, timeline view analysis, and bibliographic coupling analysis are conducted to disclose the current challenges and future research directions. This paper contributes as a remarkable reference for invaluable further research of concept drift, which enlightens the emerging/trend topics, and the possible research directions with several graphs, visualized by using the VOS viewer and Cite Space software. [ABSTRACT FROM AUTHOR]

Learning in non-stationary environments remains challenging due to dynamic and unknown probability distribution. This issue is even more problematic when there is a lack of supervision data for a specific domain, making the use of labeled data from a related but different domain highly valuable. This paper addresses the streaming data classification and introduces a heterogeneous unsupervised domain adaptation method. To cover the uncertainty caused by the distribution discrepancy and concept drifting data, the proposed method prioritizes target domain data with the highest uncertainty, as they indicate changes in data distribution. It utilizes a fuzzy-based feature-level adaptation and optimizes parameters through accelerated optimization. Additionally, it employs instance selection in the source domain to identify qualified samples, further enhancing classification and adaptation. Three different settings of the proposed method have been configured, and five state-of-the-art methods have been selected as competing methods. Regarding different types of concept drift, various experiments taken from four benchmark datasets demonstrate the superiority of the proposed method in terms of accuracy and computational time. The Wilcoxon statistical test has been conducted to prove a meaningful distinction between the evaluation metrics results of the proposed method and the competing ones. [ABSTRACT FROM AUTHOR]

Social networks have become an essential part of our lives today, at least in their virtual dimension, and the image of the web world is almost impossible without the presence of this pervasive phenomenon. These networks are one of the important components of the information infrastructure, such as twitter networks, facebook networks, and so on. In the analysis of social networks, one of the important issues is the detection of community. Each community is a group of network nodes so that the connection between nodes within the group with each other is more than their connection with other network nodes. Various methods have been proposed for community detection. One of the existing methods is based on data stream clustering. The output data of a social network can be modeled with a data stream. Fast and accurate clustering of this data stream can be very effective in the detection of community. In this research, using a fast and accurate online clustering algorithm, the community is detected. The simulation results indicate that the method proposed in this research can calculate the number of clusters optimally and perform better than similar methods. The proposed algorithm can be used in many other applications. [ABSTRACT FROM AUTHOR]

The adoption of weighted regularized extreme learning machines (WR-ELMs) has been recognized as an effective approach to addressing class imbalance by differentially weighting sample classes. Traditional batch learning methodologies, however, falter due to their inefficiency in adapting to network restructuring and their inability to process streaming data. By introducing incremental and sequential learning, the incremental weighted regularized extreme learning machine (IWR-ELM) and the online weighted regularized extreme learning machine (OWR-ELM) are proposed in this paper to enhance WR-ELM's flexibility and responsiveness. Specifically, the IWR-ELM facilitates optimal hidden layer node selection, thereby enhancing model adaptability without necessitating full retraining. Conversely, the OWR-ELM is engineered for real-time data stream processing, enabling continuous learning from new data segments without retaining outdated information. We also address the concurrent challenges of concept drift and class imbalance by presenting an enhanced online weighted regularized extreme learning machine, which incorporates enhancement factors to elevate the significance of recent data. Finally, the competitiveness of our proposed algorithm is demonstrated in terms of its training time and performance through extensive experiments conducted on class-imbalanced datasets. Our comprehensive evaluations on diverse class-imbalanced datasets affirm the superior efficiency and performance of our proposed solutions in terms of training speed and accuracy. [ABSTRACT FROM AUTHOR]

The data stream is a dynamic collection of data that changes over time, and predicting the data class can be challenging due to sparse samples, complex interdependent characteristics between data, and random fluctuations. Accurately predicting the data stream in sparse data can create complex challenges. Due to its incremental learning nature, the neural networks suitable approach for streaming visualization. However, the high computational cost limits their applicability to high-speed streams, which has not yet been fully explored in the existing approaches. To solve these problems, this paper proposes an end-to-end dynamic separation neural network (DSN) approach based on the characteristics of data stream fluctuations, which expands the static sample at a given moment into a sequence of sample streams in the time dimension, thereby increasing the sparse samples. The Temporal Augmentation Module (TAM) can overcome these challenges by modifying the sparse data stream and reducing time complexity. Moreover, a neural network that uses a Variance Detection Module (VDM) can effectively detect the variance of the input data stream through the network and dynamically adjust the degree of differentiation between samples to enhance the accuracy of forecasts. The proposed method adds significant information regarding the data sparse samples and enhances low dimensional samples to high data samples to overcome the sparse data stream problem. In VDM the preprocessed data achieve data augmentation and the samples are transmitted to VDM. The proposed method is evaluated using different types of data streaming datasets to predict the sparse data stream. Experimental results demonstrate that the proposed method achieves a high prediction accuracy and that the data stream has significant effects and strong robustness compared to other existing approaches. [ABSTRACT FROM AUTHOR]

Data stream classification is a challenging task because of disruptive changes in the data distribution, also known as concept drift. Ensemble diversification is a crucial method in data stream classification, offering improved adaptability, flexibility, and efficiency. In such cases, it is recognized that having an additional diverse ensemble of components improves prediction accuracy. Existing works have shown serious drawbacks in terms of accuracy and response time. This requires an adaptive approach for selecting components with high performance. Therefore, in this paper, we proposed an incremental ensemble diversification approach in data streams classification based on the combination of Improved Hoeffding Trees and Thompson Sampling (IHTTS). Our proposed approach begins with generating an initial set of classes for the data stream with timestamp (tn), then updating the classes when newly incoming data arrive (tn+1), and finally combining module diversity and prediction accuracy. The results on real datasets verify the efficiency and effectiveness of the proposed IHTTS approach. [ABSTRACT FROM AUTHOR]

Supervised learning algorithms generally assume the availability of enough memory to store data models during the training and test phases. However, this assumption is unrealistic when data comes in the form of infinite data streams, or when learning algorithms are deployed on devices with reduced amounts of memory. In this paper, we adapt the online Mondrian forest classification algorithm to work with memory constraints on data streams. In particular, we design five out-of-memory strategies to update Mondrian trees with new data points when the memory limit is reached. Moreover, we design node trimming mechanisms to make Mondrian trees more robust to concept drifts under memory constraints. We evaluate our algorithms on a variety of real and simulated datasets, and we conclude with recommendations on their use in different situations: the Extend Node strategy appears as the best out-of-memory strategy in all configurations, whereas different node trimming mechanisms should be adopted depending on whether a concept drift is expected. All our methods are implemented in the OrpailleCC open-source library and are ready to be used on embedded systems and connected objects. [ABSTRACT FROM AUTHOR]

Abstract The detection of different types of concept drift has wide applications in the fields of cloud computing and security information detection. Concept drift detection can indeed assist in promptly identifying instances where model performance deteriorates or when there are changes in data distribution. This paper focuses on the problem of concept drift detection in order to conduct frequent pattern mining. To address the limitation of fixed sliding windows in adapting to evolving data streams, we propose a variable sliding window frequent pattern mining algorithm, which dynamically adjusts the window size to adapt to new concept drifts and detect them in a timely manner. Furthermore, considering the challenge of existing concept drift detection algorithms that struggle to adapt to different types of drifting data simultaneously, we introduce an additional dual-layer embedded variable sliding window. This approach helps differentiate types of concept drift and incorporates a decay model for drift adaptation. The proposed algorithm can effectively detect different types of concept drift in data streams, perform targeted drift adaptation, and exhibit efficiency in terms of time complexity and memory consumption. Additionally, the algorithm maintains stable performance, avoiding abrupt changes due to window size variations and ensuring overall robustness.

The proliferation of high-dimensional data in many advanced database applications is a result of today’s technological advancements. These data points that correspond to objects are often without a precise description, which make their representation uncertain. While the concept of data streaming is not new, its practical uses are only recently emerging. This research focuses on continuous range data—a type of uncertain data common in database applications—that do not have explicit representations of their exact values. Furthermore, the identification of skyline objects—one of the popular database applications—becomes more challenging when skylines are to be identified from a collection of continuously generated input data streams where objects might have imprecise values. This makes it imperative to determine which approach has the optimal accuracy for estimating or predicting the uncertain values and at the same time able to handle a massive streams of data that are continuously generated and analyze them almost instantly to provide accurate and timely responses. Given this, the following techniques are selected—Linear Regression (LR), k-Nearest Neighbour (k-NN), Random Forest (RF), Decision Trees (DT), and Centre and Range Method (CRM) and their effectiveness is evaluated in terms of execution time, precision, recall, F1-score, and root mean square error (RMSE). Additionally, in order to verify the accuracy of each prediction technique, the predicted data derived from its model is used to derive skyline objects, which are subsequently compared to the actual skyline results. An inaccurate prediction of a continuous range value would result in incorrect set of skyline objects.

Data stream is a major data type in modern information equipment such as sensory and video/sound devices, which is fast and continuously generated without any stall cycle. Due to artificial intelligence becoming a common tool to extend application domain for information equipment, reduction of streaming data via communication path is an emerged and crucial theme targeted to support fast and large data migration among edge devices and cloud servers for IoT applications. This paper proposes a novel lossless compression algorithm based on digram coding scheme called universal Adaptive Stream-based Entropy (ASE) coding. It handles data stream with an instant entropy calculation according to table occupation ratio and provides a universal compression mechanism by matching multiple data units in data stream. It organizes a compression pipeline of multiple ASE coding modules that scans a data stream and performs encoding from multiple data units into at least a single bit of a compressed data without pipeline stall. Universal ASE coding is suitable for hardware implementation with configurable pipeline depth and resource size of the comp/decompression module. This paper describes organization of the pipeline and the control mechanisms of the comp/decompression. The evaluation reveals effective compression performances according to the effective compression mechanism and optimization techniques.

As machine learning models are increasingly applied to real-world scenarios, it is essential to consider the possibility of changes in the data distribution over time. Concept drift detection and adaptation refers to the process of identifying and tracking these changes and updating the model accordingly. Researchers have devoted significant efforts to develop various techniques and tools for concept drift detection and adaptation, as this paper provides a generic roadmap and review of the field. In this paper, we begin by reviewing the background of data stream classification and its assumptions and requirements. Then, we explore the historical development of concept drift detection and adaptation and highlight the key points of approaches that have emerged over time. Next, we summarize the major findings, challenges, and limitations of past research, and provide insights into potential future directions of the field. The paper can benefit researchers and practitioners who seek to navigate the challenges and opportunities in concept drift detection and adaptation.

Many applications and fields produce a vast quantity of time-relevant or continuously changing data which may represent new phenomena. This data stream behavior is known as Concept Drift. The need to efficiently and accurately process online data streams is a current need in many areas. Concept drift is a cause of performance degradation of classical machine learning approaches. It is necessary to address the concept drift to deploy real-world applications fed by data streams. This work presents a perspective of Concept Drift Detector (CDD) application to empower a data stream classifier in a real-world scenario followed by the proposal of Concept Drift Detector based on Typicality and Eccentricity Data Analytics (TEDA-CDD). Our method employs two models in monitoring the data stream in order to keep the information of a previous concept whereas monitoring the emergence of a new concept. The models are considered to represent two distinct concepts when the intersection of data samples are significantly low, described by the Jaccard Index. TEDA-CDD is compared to known methods from literature in experiments using synthetic and real-world datasets simulating real-world applications. In these experiments, TEDA-CDD performs comparably in terms of accuracy against well-established algorithms whereas presenting higher memory efficiency.

The development in the fields of clean energy, particularly wind and photovoltaic power, generates a large amount of data streams, and how to mine valuable information from these data to improve the efficiency of power generation has become a hot spot of current research. Traditional classification algorithms cannot cope with dynamically changing data streams, so data stream classification techniques are particularly important. The current data stream classification techniques mainly include decision trees, neural networks, Bayesian networks, and other methods, which have been applied to wind power and photovoltaic power data processing in existing research. However, the data drift problem is gradually highlighted due to the dynamic change in data, which significantly impacts the performance of classification algorithms. This paper reviews the latest research on data stream classification technology in wind power and photovoltaic applications. It provides a detailed introduction to the data drift problem in machine learning, which significantly affects algorithm performance. The discussion covers covariate drift, prior probability drift, and concept drift, analyzing their potential impact on the practical deployment of data stream classification methods in wind and photovoltaic power sectors. Finally, by analyzing examples for addressing data drift in energy-system data stream classification, the article highlights the future prospects of data drift research in this field and suggests areas for improvement. Combined with the systematic knowledge of data stream classification techniques and data drift handling presented, it offers valuable insights for future research.

The detection of different types of concept drift has wide applications in the fields of cloud computing and security information detection. Concept drift detection can indeed assist in promptly identifying instances where model performance deteriorates or when there are changes in data distribution. This paper focuses on the problem of concept drift detection in order to conduct frequent pattern mining. To address the limitation of fixed sliding windows in adapting to evolving data streams, we propose a variable sliding window frequent pattern mining algorithm, which dynamically adjusts the window size to adapt to new concept drifts and detect them in a timely manner. Furthermore, considering the challenge of existing concept drift detection algorithms that struggle to adapt to different types of drifting data simultaneously, we introduce an additional dual-layer embedded variable sliding window. This approach helps differentiate types of concept drift and incorporates a decay model for drift adaptation. The proposed algorithm can effectively detect different types of concept drift in data streams, perform targeted drift adaptation, and exhibit efficiency in terms of time complexity and memory consumption. Additionally, the algorithm maintains stable performance, avoiding abrupt changes due to window size variations and ensuring overall robustness. [ABSTRACT FROM AUTHOR]

In many industrial applications, obtaining labeled observations is not straightforward as it often requires the intervention of human experts or the use of expensive testing equipment. In these circumstances, active learning can be highly beneficial in suggesting the most informative data points to be used when fitting a model. Reducing the number of observations needed for model development alleviates both the computational burden required for training and the operational expenses related to labeling. Online active learning, in particular, is useful in high‐volume production processes where the decision about the acquisition of the label for a data point needs to be taken within an extremely short time frame. However, despite the recent efforts to develop online active learning strategies, the behavior of these methods in the presence of outliers has not been thoroughly examined. In this work, we investigate the performance of online active linear regression in contaminated data streams. Our study shows that the currently available query strategies are prone to sample outliers, whose inclusion in the training set eventually degrades the predictive performance of the models. To address this issue, we propose a solution that bounds the search area of a conditional D‐optimal algorithm and uses a robust estimator. Our approach strikes a balance between exploring unseen regions of the input space and protecting against outliers. Through numerical simulations, we show that the proposed method is effective in improving the performance of online active learning in the presence of outliers, thus expanding the potential applications of this powerful tool. [ABSTRACT FROM AUTHOR]

Data stream clustering is integral to contemporary big data applications. However, addressing the ongoing influx of data streams efficiently and accurately remains a primary challenge in current research. This paper aims to elevate the efficiency and precision of data stream clustering, leveraging the TEDA (Typicality and Eccentricity Data Analysis) algorithm as a foundation, we introduce improvements by integrating a nearest neighbor search algorithm to enhance both the efficiency and accuracy of the algorithm. The original TEDA algorithm, grounded in the concept of "Typicality and Eccentricity Data Analytics", represents an evolving and recursive method that requires no prior knowledge. While the algorithm autonomously creates and merges clusters as new data arrives, its efficiency is significantly hindered by the need to traverse all existing clusters upon the arrival of further data. This work presents the NS-TEDA (Neighbor Search Based Typicality and Eccentricity Data Analysis) algorithm by incorporating a KD-Tree (K-Dimensional Tree) algorithm integrated with the Scapegoat Tree. Upon arrival, this ensures that new data points interact solely with clusters in very close proximity. This significantly enhances algorithmefficiencywhile preventing a single data point fromjoining too many clusters andmitigating the merging of clusters with high overlap to some extent. We apply the NS-TEDA algorithm to several well-known datasets, comparing its performance with other data stream clustering algorithms and the original TEDA algorithm. The results demonstrate that the proposed algorithm achieves higher accuracy, and its runtime exhibits almost linear dependence on the volume of data, making it more suitable for large-scale data stream analysis research. [ABSTRACT FROM AUTHOR]

Super spread detection has been widely applied in network management, recommender systems, and cyberspace security. It is more complicated than heavy hitter owing to the requirement of duplicate removal. Accurately detecting a super spread in real-time with small memory demands remains a nontrivial yet challenging issue. The previous work either had low accuracy or incurred heavy memory overhead and could not provide a precise cardinality estimation. This paper designed an invertible sketch for super spread detection with small memory demands and high accuracy. It introduces a power-weakening increment strategy that creates an environment encouraging sufficient competition at the early stages of discriminating a super spread and amplifying the comparative dominance to maintain accuracy. Extensive experiments have been performed based on actual Internet traffic traces and recommender system datasets. The trace-driven evaluation demonstrates that our sketch actualizes higher accuracy in super spread detection than state-of-the-art sketches. The super spread cardinality estimation error is 2.6–19.6 times lower than that of the previous algorithms. [ABSTRACT FROM AUTHOR]

The growth of network-connected devices has led to an exponential increase in data generation, creating significant challenges for efficient data analysis. This data is generated continuously, creating a dynamic flow known as a data stream. The characteristics of a data stream may change dynamically, and this change is known as concept drift. Consequently, a method for handling data streams must efficiently reduce their volume while dynamically adapting to these changing characteristics. This article proposes a simple online vector quantization method for concept drift. The proposed method identifies and replaces units with low win probability through remove-birth updating, thus achieving a rapid adaptation to concept drift. Furthermore, the results of this study show that the proposed method can generate minimal dead units even in the presence of concept drift. This study also suggests that some metrics calculated from the proposed method will be helpful for drift detection. [ABSTRACT FROM AUTHOR]

Recently, there has been an explosive generation of streaming data in various fields such as IoT and network attack detection, medical data monitoring, and financial trend analysis. These domains require precise and rapid analysis capabilities by minimizing noise from continuously generated raw data. In this paper, we propose SPinDP (Stream Purifier in Distributed Platform), an open source-based high-speed stream purification platform, to support real-time stream purification. SPinDP consists of four major components, Data Stream Processing Engine, Purification Library, Plan Manager, and Shared Storage, and operates based on open-source systems including Apache Storm and Apache Kafka. In these components, stream processing throughput and latency are critical performance metrics, and SPinDP significantly enhances distributed processing performance by utilizing the ultra-high-speed network RDMA (Remote Direct Memory Access). For the performance evaluation, we use a distributed cluster environment consisting of nine nodes, and we show that SPinDP's stream processing throughput is more than 28 times higher than that of the existing Ethernet environment. SPinDP also significantly reduces the processing latency by more than 2473 times on average. These results indicate that the proposed SPinDP is an excellent integrated platform that can efficiently purify high-speed and large-scale streams through RDMA-based distributed processing. [ABSTRACT FROM AUTHOR]

Data are generated at an unprecedented rate and scale these days across many disciplines. The field of streaming data analysis has emerged as a result of new data collection and storage technologies in various areas, such as air pollution monitoring, detection of traffic congestion, disease surveillance, and recommendation systems. In this paper, we consider the problem of model estimation for data streams in reproducing kernel Hilbert spaces. We propose an adaptive supervised learning method with a data sparsity constraint that uses limited storage spaces and can handle nonstationary models. We demonstrate the competitive performance of the proposed method using simulations and analysis of the bike sharing dataset. [ABSTRACT FROM AUTHOR]

Accurate influent flow rate prediction is important for operators and managers at wastewater treatment plants (WWTPs), as it is closely related to wastewater characteristics such as biochemical oxygen demand (BOD), total suspend solids (TSS), and pH. Previous studies have been conducted to predict influent flow rate, and it was proved that data-driven models are effective tools. However, most of these studies have focused on batch learning, which is inadequate for wastewater prediction in the era of COVID-19 as the influent pattern changed significantly. Online learning, which has distinct advantages of dealing with stream data, large data set, and changing data pattern, has a potential to address this issue. In this study, the performance of conventional batch learning models Random Forest (RF), K-Nearest Neighbors (KNN), and Multi-Layer Perceptron (MLP), and their respective online learning models Adaptive Random Forest (aRF), Adaptive K-Nearest Neighbors (aKNN), and Adaptive Multi-Layer Perceptron (aMLP), were compared for predicting influent flow rate at two Canadian WWTPs. Online learning models achieved the highest R2, the lowest MAPE, and the lowest RMSE compared to conventional batch learning models in all scenarios. The R2 values on testing data set for 24-h ahead prediction of the aRF, aKNN, and aMLP at Plant A were 0.90, 0.73, and 0.87, respectively; these values at Plant B were 0.75, 0.78, and 0.56, respectively. The proposed online learning models are effective in making reliable predictions under changing data patterns, and they are efficient in dealing with continuous and large influent data streams. They can be used to provide robust decision support for wastewater treatment and management in the changing era of COVID-19 and also under other unprecedented emergencies that could change influent patterns. [ABSTRACT FROM AUTHOR]

BackgroundArtificial intelligence (AI) and machine learning (ML) are poised to transform infectious disease testing. Uniquely, infectious disease testing is technologically diverse spaces in laboratory medicine, where multiple platforms and approaches may be required to support clinical decision-making. Despite advances in laboratory informatics, the vast array of infectious disease data is constrained by human analytical limitations. Machine learning can exploit multiple data streams, including but not limited to laboratory information and overcome human limitations to provide physicians with predictive and actionable results. As a quickly evolving area of computer science, laboratory professionals should become aware of AI/ML applications for infectious disease testing as more platforms are become commercially available.ContentIn this review we: (a) define both AI/ML, (b) provide an overview of common ML approaches used in laboratory medicine, (c) describe the current AI/ML landscape as it relates infectious disease testing, and (d) discuss the future evolution AI/ML for infectious disease testing in both laboratory and point-of-care applications.SummaryThe review provides an important educational overview of AI/ML technique in the context of infectious disease testing. This includes supervised ML approaches, which are frequently used in laboratory medicine applications including infectious diseases, such as COVID-19, sepsis, hepatitis, malaria, meningitis, Lyme disease, and tuberculosis. We also apply the concept of "data fusion" describing the future of laboratory testing where multiple data streams are integrated by AI/ML to provide actionable clinical knowledge.

Learning from changing data streams is one of the important tasks of data mining. The phenomenon of the underlying distribution of data streams changing over time is called concept drift. In classification decision-making, the occurrence of concept drift will greatly affect the classification efficiency of the original classifier, that is, the old decision-making model is not suitable for the new data environment. Therefore, dealing with concept drift from changing data streams is crucial to guarantee classifier performance. Currently, most concept drift detection methods apply the same detection strategy to different data streams, with little attention to the uniqueness of each data stream. This limits the adaptability of drift detectors to different environments. In our research, we designed a unique solution to address this issue. First, we proposed a variance estimation strategy and a variance feedback strategy to characterize the data stream’s characteristics through variance. Based on this variance, we developed personalized drift detection schemes for different data streams, thereby enhancing the adaptability of drift detection in various environments. We conducted experiments on data streams with various types of drifts. The experimental results show that our algorithm achieves the best average ranking for accuracy on the synthetic dataset, with an overall ranking 1.12 to 1.5 higher than the next-best algorithm. In comparison with algorithms using the same tests, our method improves the ranking by 3 to 3.5 for the Hoeffding test and by 1.12 to 2.25 for the McDiarmid test. In addition, they achieve a good balance between detection delay and false positive rates. Finally, our algorithm ranks higher than existing drift detection methods across the four key metrics of accuracy, CPU time, false positives, and detection delay, meeting our expectations.