Online Stream Research Articles

An Efficient Probabilistic Algorithm to Detect Periodic Patterns in Spatio-Temporal Datasets

Deriving insight from data is a challenging task for researchers and practitioners, especially when working on spatio-temporal domains. If pattern searching is involved, the complications introduced by temporal data dimensions create additional obstacles, as traditional data mining techniques are insufficient to address spatio-temporal databases (STDBs). We hereby present a new algorithm, which we refer to as F1/FP, and can be described as a probabilistic version of the Minus-F1 algorithm to look for periodic patterns. To the best of our knowledge, no previous work has compared the most cited algorithms in the literature to look for periodic patterns—namely, Apriori, MS-Apriori, FP-Growth, Max-Subpattern, and PPA. Thus, we have carried out such comparisons and then evaluated our algorithm empirically using two datasets, showcasing its ability to handle different types of periodicity and data distributions. By conducting such a comprehensive comparative analysis, we have demonstrated that our newly proposed algorithm has a smaller complexity than the existing alternatives and speeds up the performance regardless of the size of the dataset. We expect our work to contribute greatly to the mining of astronomical data and the permanently growing online streams derived from social media.

Multiview Deep Online Clustering: An Application to Online Research Topic Modeling and Recommendations

In today&#x2019;s scenario, a large number of scientific articles on various domains are being published everyday, resulting in a rapid change in the trends of research topics. Retrieving the trending topics, evaluating the trends and extracting the scope of topics could be beneficial to young researchers, which can be recommended for future scope. Publication of articles is a continuous process, and so is the evolution of topics as well as the scope. The dynamic behavior of topics can be handled by continuously updating the partitioning of incoming articles arriving in a streaming manner. This article proposes an online clustering approach utilizing the autoencoder, which is capable of handling the online stream of articles. The training is performed by considering the multiple views of articles and posing the topic modeling (TM) problem as a multiview (MV) clustering problem. In addition, we employ an evolutionary-based approach to the latent representation of data to automatically determine the number of clusters. To learn the nonlinear mapping appropriately and to generate clusters effectively, the model simultaneously optimizes the reconstruction loss and clustering loss in an MV framework. The developed method has experimented on <i>ArXiv</i> dataset to determine the trending topics, scope, and recommendation of topics. The superiority of the proposed method in generating clusters as well as in determining scope is shown by comparing the results with many existing and baseline methods.

Network attack classification framework based on Autoencoder model and online stream analysis technology

Abstract— To deal with diverse and constantly changing forms of cyberattacks, machine learning methods have been researched and applied extensively in network data processing for positive results in network attack detection. However, machine learning models require extensive computational resources and their application to handle significant real-time data flow monitoring problems still needs improvement. In this paper, we research and propose a network attack detection framework using a 2-stage classification algorithm with an Autoencoder model, integrating online stream processing technology based on Apache Kafka and Spark technology. The results show that the proposed framework has high efficiency in detecting network attacks and faster processing time than traditional data processing technology.

PANDEMIC OASIS: POPULAR RELIGIOSITIES AS WOMEN’S LIFE-GIVING COMMUNITAS

This article argues and explores how popular Catholic piety can serve as an agent to create life-giving communitas forwomen in Macao, China. This research uses the narratives of six Catholic women about how their immersion in various public and private devotional practices creates solidarity and communities that are inclusive, empowering, and nurturing during the outbreak of COVID-19. Besides petitioning their long-lasting devotional images such as St Roch and Our Lady of Fatima, women learn and move on to a new model of online and live stream religious response and ritual innovations at the time of coronavirus in Macao. The digitisation of gathering strengthens the global connection among women to the rest of the world, shapinga new form of devotional culture and community ties that are not necessarily institutionalised. Devotional practices act as a spiritual oasis, personally and communally, for women coming together and bring hope, strength, and consolation to this unprecedented time.

Influence of 4th generation broadband cellular network technology on undergraduate education

The influence of 4th generation broadband cellular network technology on education is explored in this paper. With the widespread use of smartphones and tablets, students are increasingly using mobile devices to access learning resources, communicate with peers, and complete assignments. The advent of 4G technology has had a revolutionary impact on undergraduate education, transforming the way students access information and interact with course materials. With 4G, students can now quickly and easily access online resources, download and stream multimedia content, collaborate with classmates in real-time, and participate in remote learning. This technology has had a profound influence on the way education is managed today, providing faster and more efficient access to educational resources and a personalized learning experience while streamlining administrative processes and assisting teachers in instruction. 4G technology has completely revolutionized the way education is thought of and experienced, creating an open environment, interactivity, and increasingly specialized training for students and professionals alike.

Neo-Gnosticism and the American Dissident Right

Neo-Gnosticism and the American Dissident Right

Tuned Compositional Feature Replays for Efficient Stream Learning.

Our brains extract durable, generalizable knowledge from transient experiences of the world. Artificial neural networks come nowhere close to this ability. When tasked with learning to classify objects by training on nonrepeating video frames in temporal order (online stream learning), models that learn well from shuffled datasets catastrophically forget old knowledge upon learning new stimuli. We propose a new continual learning algorithm, compositional replay using memory blocks (CRUMB), which mitigates forgetting by replaying feature maps reconstructed by combining generic parts. CRUMB concatenates trainable and reusable memory block vectors to compositionally reconstruct feature map tensors in convolutional neural networks (CNNs). Storing the indices of memory blocks used to reconstruct new stimuli enables memories of the stimuli to be replayed during later tasks. This reconstruction mechanism also primes the neural network to minimize catastrophic forgetting by biasing it toward attending to information about object shapes more than information about image textures and stabilizes the network during stream learning by providing a shared feature-level basis for all training examples. These properties allow CRUMB to outperform an otherwise identical algorithm that stores and replays raw images while occupying only 3.6% as much memory. We stress-tested CRUMB alongside 13 competing methods on seven challenging datasets. To address the limited number of existing online stream learning datasets, we introduce two new benchmarks by adapting existing datasets for stream learning. With only 3.7%-4.1% as much memory and 15%-43% as much runtime, CRUMB mitigates catastrophic forgetting more effectively than the state-of-the-art. Our code is available at https://github.com/MorganBDT/crumb.git.

Searching for and blocking destructive content in images

Children and underage teenagers use the Internet not only for communication. So, for 2020, children aged five to seven years in half of cases use the network to play online games, at the age of eight to eleven years, this figure increases to 78%, and underage teenagers play online games in 80% of cases. According to the Ofcom article "Children and parents: media use and attitudes report 2021/22", already 30% of children under seven use messengers and social media applications, 93% use online video platforms such as YouTube, 33% watch online streams, and 88% watch TV and use the video on demand system. The older the children get, the greater the percentage of them use the Internet. Thus, the issues of protecting children from destructive information from the Internet are relevant. The article provides an analytical review of modern software solutions for the protection of children and adolescents such as SafeKiddo Parental Control, mSpy, Kids Place, KidLogger. In addition, the description of the developed software product is carried out, which allows detecting destructive content that can negatively affect children, as well as introducing new prohibited links and words.

The loss of serving in the dark

We study the following balls and bins stochastic process: There is a buffer with B bins, and there is a stream of balls X=〈X1,X2,…,XT〉 such that Xi is the number of balls that arrive before time i but after time i−1. Once a ball arrives, it is stored in one of the unoccupied bins. If all the bins are occupied then the ball is thrown away. In each time step, we select a bin uniformly at random, clear it, and gain its content. Once the stream of balls ends, all the remaining balls in the buffer are cleared and added to our gain. We are interested in analyzing the expected gain of this randomized process with respect to that of an optimal gain-maximizing strategy, which gets the same online stream of balls, and clears a ball from a bin, if exists, at any step. We name this gain ratio the loss of serving in the dark.In this paper, we determine the exact loss of serving in the dark. We prove that the expected gain of the randomized process is worse by a factor of ρ+ϵ from that of the optimal gain-maximizing strategy where ϵ=O(1B1/3) and ρ=maxα>1⁡αeα/((α−1)eα+e−1)≈1.69996. We also demonstrate that this bound is essentially tight as there are specific ball streams for which the above-mentioned gain ratio tends to ρ. Our stochastic process occurs naturally in packets scheduling and mechanisms design applications.

Automated Classification of Sleep Stages Using Single-Channel EEG

The main contribution of this paper is to present a novel approach for classifying the sleep stages based on optimal feature selection with ensemble learning stacking model using single-channel EEG signals.To find the suitable features from extracted feature vector, we obtained the ReliefF (ReF), Fisher Score (FS) and Online Stream Feature Selection (OSFS) selection algorithms.The proposed research work was performed on two different subgroups of sleep data of ISRUC-Sleep dataset. The experimental results of the proposed methodology signify that single-channel of EEG signal superior to other machine learning classification models with overall accuracies of 97.93%, 97%, and 95.96% using ISRUC-Sleep subgroup-I (SG-I) data and similarly the proposed model achieved an overall accuracies of 98.16%, 98.78%, and 95.26% using ISRUC-Sleep subgroup-III (SG-III) data with FS, ReF and OSFS respectively.

SHWS: Stochastic Hybrid Workflows Dynamic Scheduling in Cloud Container Services

Cloud container services as the new norm of cloud resource provisioning are more flexible and widely used in workflows scheduling. However, it is challenging to minimize the cost for workflows scheduling in cloud container services, especially when workflows arrive time and tasks execution time are uncertain. In this article, a stochastic hybrid workflows [i.e., off-line batch workflows (DIWs) and online stream workflows (DSWs)] scheduling problem in cloud container services is solved. A stochastic hybrid workflows scheduling system (SHWS), which consists of a workflow analyzer, workflow classifier, runtime estimator, workflow scheduler, and resource manager, is designed to manage and schedule DIWs and DSWs. Based on the SHWS, a stochastic hybrid workflows scheduling algorithm (SHWSA) is proposed to minimize the cost and improve resource utilization. We conduct the experiments using both synthetic data and real-world data to evaluate the proposed SHWSA approach. The results demonstrate the superiority of SHWSA compared with the state-of-the-art algorithms. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This article investigates a stochastic hybrid workflows scheduling problem in cloud container services. We propose a stochastic hybrid workflows scheduling algorithm, which is named SHWSA. The SHWSA is designed to jointly schedule off-line batch workflows (DIWs) and online stream workflows (DSWs) for minimizing the cost and improving resource utilization in cloud container services. The basic idea is to assign tasks subdeadlines and prioritize tasks for guaranteeing workflows deadlines constraints and the processing dependence requirements of tasks. Experiments show that SHWSA outperforms some state-of-the-art algorithms.

Adaptive bitrate streaming in multi-user downlink NOMA edge caching systems with imperfect SIC

Recently, mobile networks have witnessed an increasing dominance of video traffic streams to simultaneously distribute real-time contents to massive number of user devices. In this context, fifth-generation (5G) technologies and beyond expose their advanced new radio access interfaces to facilitate multi-user associations with successive interference cancellation (SIC)-based non-orthogonal multiple access (NOMA) mechanisms. However, the imperfectness of the SIC in reality may result in significant performance decrease in NOMA owing to inter-channel interference existence. Motivated by this observation, this paper studies adaptive bitrate streaming services given such a challenging transmission environment assumed. In this context, our objectives are to maximize the video bitrate (accordingly the video resolution) of the online streams while retaining the playback smoothness. The problem is transformed into a drift-plus-penalty balancing optimization, which is then resolved by an approximation algorithm. Numerical results highlight the outperformance of the proposed approach compared to existing algorithms in terms of video quality and smoothness in various system scenarios.

Improving Pedestrian Prediction Models With Self-Supervised Continual Learning

Autonomous mobile robots require accurate human motion predictions to safely and efficiently navigate among pedestrians, whose behavior may adapt to environmental changes. This letter introduces a self-supervised continual learning framework to improve data-driven pedestrian prediction models online across various scenarios continuously. In particular, we exploit online streams of pedestrian data, commonly available from the robot&#x2019;s detection and tracking pipeline, to refine the prediction model and its performance in unseen scenarios. To avoid the forgetting of previously learned concepts, a problem known as catastrophic forgetting, our framework includes a regularization loss to penalize changes of model parameters that are important for previous scenarios and retrains on a set of previous examples to retain past knowledge. Experimental results on real and simulation data show that our approach can improve prediction performance in unseen scenarios while retaining knowledge from seen scenarios when compared to naively training the prediction model online.

An in-depth study of continuous subgraph matching

Continuous subgraph matching (CSM) algorithms find the occurrences of a given pattern on a stream of data graphs online. A number of incremental CSM algorithms have been proposed. However, a systematical study on these algorithms is missing to identify their advantages and disadvantages on a wide range of workloads. Therefore, we first propose to model CSM as incremental view maintenance (IVM) to capture the design space of existing algorithms. Then, we implement six representative CSM algorithms, including InclsoMatch, SJ-Tree, Graphflow, IEDyn, TurboFlux, and SymBi, in a common framework based on IVM. We further conduct extensive experiments to evaluate the overall performance of competing algorithms as well as study the effectiveness of individual techniques to pinpoint the key factors leading to the performance differences. We obtain the following new insights into the performance: (1) existing algorithms start the search from an edge in the query graph that maps to an updated data edge, potentially leading to many invalid partial results; (2) all matching orders are based on simple heuristics, which appear ineffective at times; (3) index updates dominate the query time on some queries; and (4) the algorithm with constant delay enumeration bears significant index update cost. Consequently, no algorithm dominate the others in all cases. Therefore, we give a few recommendations based on our experiment results. In particular, the SymBi index is useful for sparse queries or long running queries. The matching orders of IEDyn and TurboFlux work well on tree queries, those of Graphflow on dense queries or when both query and data graphs are sparse, and otherwise, we recommend SymBi's matching orders.

Online continual learning in image classification: An empirical survey

Online continual learning for image classification studies the problem of learning to classify images from an online stream of data and tasks, where tasks may include new classes (class incremental) or data nonstationarity (domain incremental). One of the key challenges of continual learning is to avoid catastrophic forgetting (CF), i.e., forgetting old tasks in the presence of more recent tasks. Over the past few years, a large range of methods and tricks have been introduced to address the continual learning problem, but many have not been fairly and systematically compared under a variety of realistic and practical settings.To better understand the relative advantages of various approaches and the settings where they work best, this survey aims to (1) compare state-of-the-art methods such as Maximally Interfered Retrieval (MIR), iCARL, and GDumb (a very strong baseline) and determine which works best at different memory and data settings as well as better understand the key source of CF; (2) determine if the best online class incremental methods are also competitive in the domain incremental setting; and (3) evaluate the performance of 7 simple but effective tricks such as the ”review” trick and the nearest class mean (NCM) classifier to assess their relative impact. Regarding (1), we observe that iCaRL remains competitive when the memory buffer is small; GDumb outperforms many recently proposed methods in medium-size datasets and MIR performs the best in larger-scale datasets. For (2), we note that GDumb performs quite poorly while MIR – already competitive for (1) – is also strongly competitive in this very different (but important) continual learning setting. Overall, this allows us to conclude that MIR is overall a strong and versatile online continual learning method across a wide variety of settings. Finally for (3), we find that all tricks are beneficial, and when augmented with the “review” trick and NCM classifier, MIR produces performance levels that bring online continual learning much closer to its ultimate goal of matching offline training. Our codes are available at https://github.com/RaptorMai/online-continual-learning.

EVALUATION CHARACTERISTICS OF THE BAGASSE DRYING PROCESS

In the process of sugar production, much attention is paid to production automation, which controls the relevant technological parameters. Drying of pulp is an expensive and energy-intensive process. The use of only drying laboratory cabinets to control the moisture content of dry pulp is time consuming and can lead to deviations in the moisture content of the product. Installing humidity sensors as elements of the automation system at facilities with different temperature regimes at the inlet and outlet of the dryer is a simple and reliable way to save money, as it allows you to measure the humidity of dry pulp in the online stream, drying temperature and gas consumption. You can adjust the temperature, drying time, speed of movement of the product through the dryer and guarantee the required level of humidity of the product for granulation and storage. The case of dependence of pulp production on gas supply is considered. The input data is the gas flow rate over a fixed time t during which the specified output is produced. Build a graph of the frequency of distribution of each random variable, grouped into intervals. There are point estimates of this sample: mathematical expectation, standard deviation of each random variable to determine the optimal use of resources. To study the dependence of random variables, the parameters of the normal distribution law for these values ​​are determined. To effectively use the pulp drying process, the dependence on the amount of gas supply to the number of products is considered. According to the given normal law of distribution of technological parameters of drying and production of pulp at the corresponding given expenses of gas it is possible to put forward a hypothesis about an interval estimation of distribution between production and consumption of gas. With a stable gas supply of 0.108, the production of pulp will be optimal. The directions of further research of the sugar production process in order to develop current mathematical models and systems for its automation are determined.

Feature interaction based online streaming feature selection via buffer mechanism

AbstractFeature selection is a nontrivial preprocessing technique in many practical application domains. There are three key challenges with respect to real‐world data. Firstly, the dimensionality of data keeps growing and will achieve hundreds of millions. Secondly, the data has the characteristic of high‐dimensional and small‐size. Thirdly, practical applications need to process each feature in an online manner. However, most of the previous methods only pay much attention to solving the challenges of high dimensionality and online stream. To address all issues above, we propose OFSI, in this article, an nline streaming eature election based on feature nteraction method for feature selection. OFSI can effectively select the streaming features that are strongly related to each other in high‐dimensional and small‐size data, via using feature interaction. Furthermore, to address upcoming features that arrive by groups, we present a new group‐OFSI algorithm for online group feature selection. An extensive experiment using a series of benchmark data sets shows that the proposed two algorithms, OFSI and group‐OFSI, outperform six state‐of‐the‐art online streaming feature selection methods.

Analysis of error-based machine learning algorithms in network anomaly detection and categorization

Intrusion and anomaly detection are particularly important to protect computer networks and communication vulnerability. This research aims to experimentally identify the best error-based machine learning algorithm for anomaly detection and anomaly attack categorization with the highest accuracy and fastest build time. A two-stage anomaly and categorization framework has been set up for experimental evaluation. The first stage identifies if a network flow is normal or anomalous and the second stage identifies type of attack if the first stage result is anomalous. The goal is to eventually use the best algorithm in an online stream model of network intrusion detection. To this end, five research propositions are defined, four sets of experiments are set up, and four research questions are asked. The UNSW-NB15 dataset for network anomaly is used for training and testing in the experiments. Machine learning algorithms are classified into four different learning approaches: information-based, similarity-based, probability-based, and error-based. Our focus in this paper is on the error-based learning models, specifically, the following algorithms: Winnow, Logistic, Perceptron, Support Vector Machine (SVM), and Deep Learning. The results are also compared with the results of non-error-based machine learning algorithms. The results obtained show that, overall, the error-based machine learning algorithm, Winnow, is the best with 100% accuracy and time to build the model of 0.47 s for network anomaly detection. In terms of accuracy only, SVM comes top for network anomaly attack categorization but Simple Logistic is the best when accuracy and time to build are considered together.

Implementing Flexible Systems in Doctoral Viva Defense Through Virtual Mechanism.

Year 2020 has cultivated new arena in education. Covid-19 has not only reformed the world but also revived the process of operations. One among them is academics. Never in history of education was a lecture delivered without the physical presence of lecturers, in an academic environment. Even in present wildest innovative decade, the idea of eliminating the academia was not an option. However, after the pandemic encounter, it happened. These are not mere changes, but it is a transformation. While this phenomenon is getting acceptable gradually by takers, the mammoth turnaround is happening in Ph.D. final viva voce defenses too. Practice of online defense is approved by leading Universities so as to avoid stagnation of researchers in final stage. Question is, will it be appropriate to accept final defenses in online stream? and has the institutions developed rubrics to assess the researches which are presented online? Finally, if this is the only choice then how best we can use the technology to generalize it as a system? This research was executed through Delphi technique method from academia, research aspirants and policy-makers. Partial least square is used to quantify the data after interpreting qualitative data. Thirteen homogeneous University’s respondents were used as sample. Focus of this paper remains to substantiate the benefits of emerging technology and developing a scheme for virtual streaming, rather identifying reasons to avoid them.

НЕОИМПЕРСКИЙ ЛЕЙТМОТИВ В ИСТОРИЧЕСКИХ РОССИЙСКИХ ФИЛЬМАХ 2000–2010-х гг. И ФЕНОМЕН КИНЕМАТОГРАФИЧЕСКОЙ ПРАВОСЛАВНОЙ ПРОПОВЕДИ

During the last two decades, the topic of imperial consciousness and the ideological construction of a neo-imperial approach to contemporary Russian history within the narrative of post-Soviet “neo-imperialism” has been increasingly revived in Russian cinema. It has not much about developing the idea of Great Russia with a highest world mission. There are hardly more Slavophile motives here than Westernisers’ ones. The new historical movies shot mainly by the Russian state or Church request construct a parable about a great supranational historical empire, which unites all ethnic groups, peoples, and religions under the rule of the Russian Tsar / Emperor / Patriarch. For Soviet times, the Communist Party or its leader usually plays the role of the monarch, while the role of civil religion can be played by political ideology. The modern Patriarchate of the Russian Orthodox Church disseminates the idea of imperialism as a basis of the Orthodox media cinematographic preaching. This type of preaching, along with Christian preaching on the Internet, social networks, and video games, has become increasingly used by Orthodox clergymen and church organisations. Here Russian Orthodox Church follows the Roman Catholics and Protestants, who currently use media sermon as an important part of their homiletics. From priests to clergy of the highest rank, such as Metropolitan Hilarion (Alfeev) of Volokolamsk or Metropolitan Tikhon (Shevkunov) of Pskov and Porkhovsk, more and more men of cloth begin to attract audience by their performance in offline and online video blogs, by online streams of their sermons in the digital space and finally, by shooting films, both documentaries and feature films. The main components of the neo-imperial narrative in the modern ideologised cinematographic preaching are the following: 1) constructing the parable of the great supranational historical empire (pre-Mongol Rus, Horde, Russian Principality, Russian Tsardom, Russian Empire, finally, the Soviet Union); 2) emphasising the great role of political leader, prince / tsar / emperor / general secretary / president; 3) resorting to preposterous synthesis of incompatible ideals of the Russian Empire and the Soviet Union; 4) spreading anti-historicism, anachronism, including linguistic, substituting historical fantasy for the history of the Russian Orthodox Church; 5) elevating the current national ideology to the status of national archetype, with searching and successful finding of “eternal” geopolitical enemies and friends of Russia; 6) re-considering Orthodoxy as a glamour “export product” for the West. The Cinematographic Orthodoxy depicted in imperial propaganda films has little in common with the history of the Russian Orthodox Church and the ideal of Orthodox Christian morality. Under the pen of scriptwriters and film directors who create their mass products by the order of the current Russian state or Patriarchate, Orthodox Christianity becomes a social and media construct that must exists solely for the blessing and support of the state power, as well as for the entertainment of the mass audience. Therefore, we must stress that such a preaching does not achieve its goals, as it does not preach Jesus Christ’s teaching, spirituality and a way for soul salvation and improvement.