Dynamic Data Streams Research Articles

Robust Online Support Vector Regression with Truncated ε-Insensitive Pinball Loss

Advances in information technology have led to the proliferation of data in the fields of finance, energy, and economics. Unforeseen elements can cause data to be contaminated by noise and outliers. In this study, a robust online support vector regression algorithm based on a non-convex asymmetric loss function is developed to handle the regression of noisy dynamic data streams. Inspired by pinball loss, a truncated ε-insensitive pinball loss (TIPL) is proposed to solve the problems caused by heavy noise and outliers. A TIPL-based online support vector regression algorithm (TIPOSVR) is constructed under the regularization framework, and the online gradient descent algorithm is implemented to execute it. Experiments are performed using synthetic datasets, UCI datasets, and real datasets. The results of the investigation show that in the majority of cases, the proposed algorithm is comparable, or even superior, to the comparison algorithms in terms of accuracy and robustness on datasets with different types of noise.

Personalized online ensemble machine learning with applications for dynamic data streams.

In this work we introduce the personalized online super learner (POSL), an online personalizable ensemble machine learning algorithm for streaming data. POSL optimizes predictions with respect to baseline covariates, so personalization can vary from completely individualized, that is, optimization with respect to subject ID, to many individuals, that is, optimization with respect to common baseline covariates. As an online algorithm, POSL learns in real time. As a super learner, POSL is grounded in statistical optimality theory and can leverage a diversity of candidate algorithms, including online algorithms with different training and update times, fixed/offline algorithms that are not updated during POSL's fitting procedure, pooled algorithms that learn from many individuals' time series, and individualized algorithms that learn from within a single time series. POSL's ensembling of the candidates can depend on the amount of data collected, the stationarity of the time series, and the mutual characteristics of a group of time series. Depending on the underlying data-generating process and the information available in the data, POSL is able to adapt to learning across samples, through time, or both. For a range of simulations that reflect realistic forecasting scenarios and in a medical application, we examine the performance of POSL relative to other current ensembling and online learning methods. We show that POSL is able to provide reliable predictions for both short and long time series, and it's able to adjust to changing data-generating environments. We further cultivate POSL's practicality by extending it to settings where time series dynamically enter and exit.

Data Stream Approach for Exploration of Droughts and Floods Driving Forces in the Dongting Lake Wetland

Wetlands are important environmental resources that are vulnerable to droughts and floods. Studying drought-flood events and their driving factors is essential for wetland resource planning and management. However, climate change and human activities present dynamic challenges that traditional approaches are unable to simulate dynamically in a rapidly changing environment. This makes quantitative analysis difficult. Our research focused on the innovative use of the data stream model, namely online bagging of Hoeffding adaptive trees, to quantify drought and flood drivers in response to climate change and human activity. The proposed approach was applied to a river-lake system, the Dongting Lake wetland. The frequency and duration characteristics of drought-flood events were analyzed. In addition, the cyclical changes of droughts and floods were analyzed by wavelet analysis. Then, drought-flood indicators as well as climatic and hydrological factors were entered into a dynamic data stream model for quantitative calculations. The results showed that the water conservancy projects largely reduced flood events while aggravating droughts. The frequency of floods decreased by 4.91% and the frequency of droughts increased by 6.81% following the construction of the Gezhouba Hydro-project and the Three Gorges Dam. Precipitation and Sankou streamflow were two dominant factors in the Dongting Lake drought and flood events, both of which had a feature importance value of approximately 0.3. This research showed how the data stream model can be used in a changing environment and the applicability of the conclusions reached through real-world instances. Moreover, these quantitative outputs can help in the sustainable utilization of Dongting Lake wetland resources.

Muscular Human Cybertwin for Internet of Everything: A Pilot Study

The cybertwin-driven 6G that can obtain static and dynamic data stream of users provide an exciting potential for a novel muscular human cybertwin beyond traditonally used artificial neural networks (ANNs) and musculoskeletal models (MSMs). In this article, we propose the conceptual design of the muscular human cybertwin and construct a baseline model with an improved generalization ability over ANN and an easier adaptation to new data distributions over MSMs. In particular, we for the first time propose to combine ANN and MSM, which benefits from the combination of learning-based approaches and analytical approaches. We then experimentally compare different manners of the combination and demonstrate the better combining manner on our testing case. Finally, we evaluate our method on an open-sourced dataset and on data from wearable sensors from the aspects of joint moment prediction accuracy, data efficiency, generalization ability, and time efficiency of personalization. Our proposed method achieves accuracy similar with ANN and over 30 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> better than MSM with sufficient training data. Compared with ANN, the improved data efficiency is presented by the better accuracies with a small amount of training data, and the generalization ability to unseen walking conditions and new subjects are demonstrated by the over 70 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> accuracy improvements. Moreover, when fine-tuning the model, our algorithm is demonstrated by the time 75 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> shorter than calibrated MSM and the accuracy improvements.

Sliding Window-Based High Utility Item-Sets Mining Over Data Stream Using Extended Global Utility Item-Sets Tree

High utility item-sets mining (HUIM) is a special topic in frequent item-sets mining (FIM). It gives better insights for business growth by focusing on the utility of items in a transaction. HUIM is evolving as a powerful research area due to its vast applications in many fields. Data stream processing, meanwhile, is an interesting and challenging problem since, processing very fast generating a huge amount of data with limited resources strongly demands high-performance algorithms. This paper presents an innovative idea to extract the high utility item-sets (HUIs) from the dynamic data stream by applying sliding window control. Even though certain algorithms exist to solve the same problem, they allow redundant processing or reprocessing of data. To overcome this, the proposed algorithm used a tree like structure called extended global utility item-sets tree (EGUI-tree), which is flexible to store and retrieve the mined information instead of reprocessing. An experimental study on real-world datasets proved that EGUI-tree algorithm is faster than the state-of-the-art algorithms.

Elastic gradient boosting decision tree with adaptive iterations for concept drift adaptation

As an excellent ensemble algorithm, Gradient Boosting Decision Tree (GBDT) has been tested extensively with static data. However, real-world applications often involve dynamic data streams, which suffer from concept drift problems where the data distribution changes overtime. The performance of GBDT model is degraded when applied to predict data streams with concept drift. Although incremental learning can help to alleviate such degrading, finding a perfect learning rate (i.e., the iteration in GBDT) that suits all time periods with all their different drift severity levels can be difficult. In this paper, we convert the issue of determining an optimal learning rate into the issue of choosing the best adaptive iterations when tuning GBDT. We theoretically prove that drift severity is closely related to the convergence rate of model. Accordingly, we propose a novel drift adaptation method, called adaptive iterations (AdIter), that automatically chooses the number of iterations for different drift severities to improve the prediction accuracy for data streams under concept drift. In a series of comprehensive tests with seven state-of-the-art drift adaptation methods on both synthetic and real-world data, AdIter yielded superior accuracy levels.

Dynamic Data Streams for Time-Critical IoT Systems in Energy-Aware IoT Devices Using Reinforcement Learning.

Thousands of energy-aware sensors have been placed for monitoring in a variety of scenarios, such as manufacturing, control systems, disaster management, flood control and so on, requiring time-critical energy-efficient solutions to extend their lifetime. This paper proposes reinforcement learning (RL) based dynamic data streams for time-critical IoT systems in energy-aware IoT devices. The designed solution employs the Q-Learning algorithm. The proposed mechanism has the potential to adjust the data transport rate based on the amount of renewable energy resources that are available, to ensure collecting reliable data while also taking into account the sensor battery lifetime. The solution was evaluated using historical data for solar radiation levels, which shows that the proposed solution can increase the amount of transmitted data up to 23%, ensuring the continuous operation of the device.

Smart pools of data with ensembles for adaptive learning in dynamic data streams with class imbalance

Streaming data incorporates dynamicity due to a nonstationary environment where data samples may endure class imbalance and change in data distribution over the period causing concept drifts. In real-life applications learning in dynamic data streams, is vitally important and challenging. A combined solution to adapt to class imbalance and concept drifts in dynamic data streams is rarely addressed by researchers. With this motivation, the current communication presents the online ensemble model smart pools of data with ensembles for class imbalance adaptive learning (SPECIAL) to learn in skewed and drifting data streams. It employs an ageing-based G-mean maximization strategy to adapt to dynamicity in data streams. It employs smart data-pools with the local expertise ensemble to classify samples lying in the same data-pool. The empirical and statistical study on different evaluation metrics exhibits that SPECIAL is more adaptive to class imbalanced dynamic data streams than the state-of-the-art algorithms.

Finding and Tracking Multi-Density Clusters in Online Dynamic Data Streams

Change is one of the biggest challenges in dynamic stream mining. From a data-mining perspective, adapting <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">and tracking</i> change is desirable in order to understand how and why change has occurred. Clustering, a form of unsupervised learning, can be used to identify the underlying patterns in a stream. Density-based clustering identifies clusters as areas of high density separated by areas of low density. This paper proposes a Multi-Density Stream Clustering (MDSC) algorithm to address these two problems; the multi-density problem and the problem of discovering and tracking changes in a dynamic stream. MDSC consists of two on-line components; discovered, labelled clusters and an outlier buffer. Incoming points are assigned to a live cluster or passed to the outlier buffer. New clusters are discovered in the buffer using an ant-inspired swarm intelligence approach. The newly discovered cluster is uniquely labelled and added to the set of live clusters. Processed data is subject to an ageing function and will disappear when it is no longer relevant. MDSC is shown to perform favourably to state-of-the-art peer stream-clustering algorithms on a range of real and synthetic data-streams. Experimental results suggest that MDSC can discover qualitatively useful patterns while being scalable and robust to noise.

ИССЛЕДОВАНИЕ ИНДИКАТОРОВ КОМПРОМЕТАЦИИ ДЛЯ СРЕДСТВ ЗАЩИТЫ ИНФОРМАЦИОННЫХ И КИБЕРФИЗИЧЕСКИХ СИСТЕМ

Purpose of work: study of existing standards of compromise indicators and methods of their exchange for enrichment of protection systems of information and cyber-physical systems. Research method: systematic analysis of open sources of data on indicators of compromise, standards of their description and methods of exchange in the organization of cyberintelligence. The result obtained: the actual problems of proactive search of threats are formulated on the example of the application of open sources of indicators of compromise in the processing of event flows in security event management systems. The classification of indicators derived from internal sources is proposed. The main problems of processing dynamic threat data streams under changing attack vectors are formulated. It was found that the threat intelligence industry currently lacks a unified solution in terms of standardization of information exchange between different platforms, but there are a number of dominant standards and formats of such data exchange. In the course of preparing the review of existing, the tasks of identifying previously unknown attack methods based on the use of open sources of indicators of compromise in data processing in security incident management systems were considered and structured, and methods for their solution were proposed. Scientific novelty: the presented article is one of the first domestic works, devoted to the analysis of research in recent years in the field of organization of work with threat intelligence data sources. Reviewed and systematized the sources of indicators of compromise and proposed their classification. Formulated the main problems of processing dynamic threat data streams under conditions of variable attack vectors.

Stream Classification Algorithm Based on Decision Tree

Due to the rise of many fields such as e-commerce platforms, a large number of stream data has emerged. The incomplete labeling problem and concept drift problem of these data pose a huge challenge to the existing stream data classification methods. In this respect, a dynamic stream data classification algorithm is proposed for the stream data. For the incomplete labeling problem, this method introduces randomization and iterative strategy based on the very fast decision tree VFDT algorithm to design an iterative integration algorithm, and the algorithm uses the previous model classification result as the next model input and implements the voting mechanism for new data classification. At the same time, the window mechanism is used to store data and calculate the data distribution characteristics in the window, then, combined with the calculated result and the predicted amount of data to adjust the size of the sliding window. Experiments show the superiority of the algorithm in classification accuracy. The aim of the study is to compare different algorithms to evaluate whether classification model adapts to the current data environment.

A multi-level collaborative framework for elastic stream computing systems

An elastic stream computing system is expected to process dynamic and volatile data streams with low latency and high throughput in timely manner. Effective management of stream application is considered one of the keys to achieve elastic computing by scaling in/out the workload of each computing node properly during runtime. Many existing work tried to build an elastic stream computing system from one perspective or at one level, which limited to some extent the system performance improvement. To address the problems brought by single level management, in this paper, we propose and implement a multi-level collaborative framework (called Mc-Stream) for elastic stream computing systems. This paper introduces our solution from the following aspects: (1) Extensive experiments show that system performance is affected by multiple factors locating at different levels. A multi-level collaborative optimization strategy can coordinate those factors and optimize the performance to a greater extent. (2) A system model is constructed to explain the multi-level collaborative framework, with the creation of topology model, data model and grouping model. The process of multi-level collaborative framework is formalized, including optimizing instances number, determining data stream load ratio among instances and deploying instances. (3) The system performance is optimized at multiple levels (user level, instance level, scheduling level, and resource level). It is further improved by the components of lightweight instances management, available resource-aware data stream redirection, fast and effective scheduling management, and asynchronous runtime redeployment without state loss. (4) Mc-Stream is implemented on top of Apache Storm platform. Metrics are evaluated with real-world stream applications, such as the fulfillment of system latency, throughput and resources utilization. Experimental results show the significant improvements made by Mc-Stream: reducing average system latency by 32%, increasing average system throughput by 26% and average resources utilization by 34%, compared with existing state-of-the-art scheduling strategies.

ADAW: Age decay accuracy weighted ensemble method for drifting data stream mining

Dynamic environment data generators are very often in real-world that produce data streams. A data source of a dynamic environment generates data streams in which the underlying data distribution changes very frequently with respect to time and hence results in concept drifts. As compared to the stationary environment, learning in the dynamic environment is very difficult due to the presence of concept drifts. Learning in dynamic environment requires evolutionary and adaptive approaches to be accommodated with the learning algorithms. Ensemble methods are commonly used to build classifiers for learning in a dynamic environment. The ensemble methods of learning are generally described at three very crucial aspects, namely, the learning and testing method employed, result integration method and forgetting mechanism for old concepts. In this paper, we propose a novel approach called Age Decay Accuracy Weighted (ADAW) ensemble architecture for learning in concept drifting data streams. The ADAW method assigned weights to the component classifiers based on its accuracy and its remaining life-time in the ensemble is such a way that ensures maximum accuracy. We empirically evaluated ADAW on benchmark artificial drifting data stream generators and real datasets and compared its performance with ten well-known state-of-the-art existing methods. The experimental results show that ADAW outperforms over the existing methods.

Dimensionality reduction in the context of dynamic social media data streams

In recent years social media became an important part of everyday life for many people. A big challenge of social media is, to find posts, that are interesting for the user. Many social networks like Twitter handle this problem with so-called hashtags. A user can label his own Tweet (post) with a hashtag, while other users can search for posts containing a specified hashtag. But what about finding posts which are not labeled by the creator? We provide a way of completing hashtags for unlabeled posts using classification on a novel real-world Twitter data stream. New posts will be created every second, thus this context fits perfectly for non-stationary data analysis. Our goal is to show, how labels (hashtags) of social media posts can be predicted by stream classifiers. In particular, we employ random projection (RP) as a preprocessing step in calculating streaming models. Also, we provide a novel real-world data set for streaming analysis called NSDQ with a comprehensive data description. We show that this dataset is a real challenge for state-of-the-art stream classifiers. While RP has been widely used and evaluated in stationary data analysis scenarios, non-stationary environments are not well analyzed. In this paper, we provide a use case of RP on real-world streaming data, especially on NSDQ dataset. We discuss why RP can be used in this scenario and how it can handle stream-specific situations like concept drift. We also provide experiments with RP on streaming data, using state-of-the-art stream classifiers like adaptive random forest and concept drift detectors. Additionally, we experimentally evaluate an online principal component analysis (PCA) approach in the same fashion as we do for RP. To obtain higher dimensional synthetic streams, we use random Fourier features (RFF) in an online manner which allows us, to increase the number of dimensions of low dimensional streams.

Research on English Movie Resource Information Mining Based on Dynamic Data Stream Classification

Massive English film and television resources will provide people with a rich content, while it has become more and more difficult for people to find the resources and information they need. Through the analysis of massive amounts of film and television data, the application system can effectively push the works that users may like. In response to this problem, we can build an English film and television resource information mining model by combining fuzzy neural network algorithms and dynamic data stream classification technology. Firstly, we use dynamic data stream classification technology to preprocess and screen English film and television resource information. Secondly, we use the fuzzy neural network algorithm to conduct data mining on related film and television resource information. The experimental results verify the superior performance of the English film and television resource information mining model established in this paper. This model can help people to find the resources and information they need.

Change detection in parametric multivariate dynamic data streams using the ARMAX-GARCH model

Dynamic data detection is one of the main concerns in the statistical process control (SPC) field. Here we focus on monitoring parametric multivariate dynamic data streams using the ARMAX-GARCH model, which reflects both the influence of exogenous variables on the mean vector and the heterogeneity of the covariance matrix. A quasi maximum likelihood estimator is used to estimate the parameter vector of a dynamic process, and a top-r control scheme is proposed to monitor the parameters of multi-dimensional data streams. Finally, a real-data example of monitoring landslide illustrates the superiorities of the proposed scheme.

Online Rebuilding Regression Random Forests

Continuous data streams mining is particularly challenging for machine learning. Many efforts have been devoted to propose online learning algorithms that can train iteratively from new coming data and provide evolutionary predictions. Compared to off-line approaches, these algorithms have shown better predictive performance and certain adaptation to high volume continuous data stream. However, a wide range of practical applications calls for regression models that can make adequate use of the large volume of pre-collected training data, meanwhile, handle continuous data stream with multi-type concept drifts, such as abrupt, gradual, incremental, recurring concept drifts. Random Forests(RFs) are an effective ensemble learning model for regression tasks. However, the fixed structure of RFs by off-line training has restricted its applicability for real-world tasks with dynamic data streams. To address these issues, we propose an online rebuilding strategy for the pre-trained Random Forests model, which is called Online Rebuilding Regression Random Forests(ORB-RRF). Specifically, a leaf-pruning technique and online reconstruction of subtrees based on the change of feature space on certain nodes are designed to adjust the local structure of regression tree to adapt to dynamic inputs. Numerical experiments with ORB-RRF show remarkable improvements in the adaptability in data stream and the predictive accuracy in several benchmark real datasets and synthetic datasets, compared to several state-of-art methods. Moreover, we show the convergence and stability of the proposed method.

Correlation Clustering in Data Streams

Clustering is a fundamental tool for analyzing large data sets. A rich body of work has been devoted to designing data-stream algorithms for the relevant optimization problems such as k-center, k-median, and k-means. Such algorithms need to be both time and and space efficient. In this paper, we address the problem of correlation clustering in the dynamic data stream model. The stream consists of updates to the edge weights of a graph on n nodes and the goal is to find a node-partition such that the end-points of negative-weight edges are typically in different clusters whereas the end-points of positive-weight edges are typically in the same cluster. We present polynomial-time, O(ncdot {{,mathrm{polylog},}}n)-space approximation algorithms for natural problems that arise. We first develop data structures based on linear sketches that allow the “quality” of a given node-partition to be measured. We then combine these data structures with convex programming and sampling techniques to solve the relevant approximation problem. Unfortunately, the standard LP and SDP formulations are not obviously solvable in O(ncdot {{,mathrm{polylog},}}n)-space. Our work presents space-efficient algorithms for the convex programming required, as well as approaches to reduce the adaptivity of the sampling.

RUE: A caching method for identifying and managing hot data by leveraging resource utilization efficiency

AbstractIn this study, we propose a caching method called RUE for dynamic large‐scale data streams. We define a data model to facilitate hot data identification and management. At the heart of RUE model is hot degree that takes into account two factors data resource utilization efficiency and reuse distance, aiming to quantitatively reflect data popularity in a dynamic data stream. Based on data's hot degree, RUE classifies data into four types, each of which is assigned with an associated cache residence time. Guided by RUE model, we develop HM algorithm to identify and manage hot data in a dynamic data stream. HM algorithm is implemented by four stacks, namely, new stack, short stack, long stack, and temp stack. Moreover, an eviction and a migration algorithms are integrated into HM to facilitate block replacement and migration. To evaluate the performance of HM algorithm, we quantitatively compare the performance of RUE with three state‐of‐art algorithms, namely, LRU, LIRS, and ARC under various replacement policies, operations, and workloads. Experimental results show that RUE outperforms these three existing algorithms in terms of both read and write hit rates. Furthermore, we show that with the four stacks in place, the computing overhead of HM is negligible.

Similarity Search for Dynamic Data Streams

Nearest neighbor searching systems are an integral part of many online applications, including but not limited to pattern recognition, plagiarism detection, and recommender systems. With increasingly larger data sets, scalability has become an important issue. Many of the most space and running time efficient algorithms are based on locality-sensitive hashing. Here, we view the data set as an $n$ n by $|U|$ | U | matrix where each row corresponds to one of $n$ n users and the columns correspond to items drawn from a universe $U$ U . The de-facto standard approach to quickly answer nearest neighbor queries on such a data set is usually a form of min-hashing. Not only is min-hashing very fast, but it is also space efficient and can be implemented in many computational models aimed at dealing with large data sets such as MapReduce and streaming. However, a significant drawback is that minhashing and related methods are only able to handle insertions to user profiles and tend to perform poorly when items may be removed. We initiate the study of scalable locality-sensitive hashing (LSH) for fully dynamic data-streams. Specifically, using the Jaccard index as similarity measure, we design (1) a collaborative filtering mechanism maintainable in dynamic data streams and (2) a sketching algorithm for similarity estimation. Our algorithms have little overhead in terms of running time compared to previous LSH approaches for the insertion only case, and drastically outperform previous algorithms in case of deletions.