Stream Data Mining Algorithms And Their Probabilistic Properties

This book presents a unique approach to stream data mining. Unlike the vast majority of previous approaches, which are largely based on heuristics, it highlights methods and algorithms that are mathematically justified. First, it describes how to adapt static decision trees to accommodate data streams; in this regard, new splitting criteria are developed to guarantee that they are asymptotically equivalent to the classical batch tree. Moreover, new decision trees are designed, leading to the original concept of hybrid trees. In turn, nonparametric techniques based on Parzen kernels and orthogonal series are employed to address concept drift in the problem of non-stationary regressions and classification in a time-varying environment. Lastly, an extremely challenging problem that involves designing ensembles and automatically choosing their sizes is described and solved. Given its scope, the book is intended for a professional audience of researchers and practitioners who deal with stream data, e.g. in telecommunication, banking, and sensor networks.

This book primarily discusses issues related to the mining aspects of data streams and it is unique in its primary focus on the subject. This volume covers mining aspects of data streams comprehensively: each contributed chapter contains a survey on the topic, the key ideas in the field for that particular topic, and future research directions. The book is intended for a professional audience composed of researchers and practitioners in industry. This book is also appropriate for advanced-level students in computer science.

This book is a significant contribution to the subject of mining time-changing data streams and addresses the design of learning algorithms for this purpose. It introduces new contributions on several different aspects of the problem, identifying research opportunities and increasing the scope for applications. It also includes an in-depth study of stream mining and a theoretical analysis of proposed methods and algorithms. The first section is concerned with the use of an adaptive sliding window algorithm (ADWIN). Since this has rigorous performance guarantees, using it in place of counters or accumulators, it offers the possibility of extending such guarantees to learning and mining algorithms not initially designed for drifting data. Testing with several methods, including Naïve Bayes, clustering, decision trees and ensemble methods, is discussed as well. The second part of the book describes a formal study of connected acyclic graphs, or 'trees', from the point of view of closure-based mining, presenting efficient algorithms for subtree testing and for mining ordered and unordered frequent closed trees. Lastly, a general methodology to identify closed patterns in a data stream is outlined. This is applied to develop an incremental method, a sliding-window based method, and a method that mines closed trees adaptively from data streams. These are used to introduce classification methods for tree data streams.

This compendium is a completely revised version of an earlier book, Data Mining in Time Series Databases, by the same editors. It provides a unique collection of new articles written by leading experts that account for the latest developments in the field of time series and data stream mining.The emerging topics covered by the book include weightless neural modeling for mining data streams, using ensemble classifiers for imbalanced and evolving data streams, document stream mining with active learning, and many more. In particular, it addresses the domain of streaming data, which has recently become one of the emerging topics in Data Science, Big Data, and related areas. Existing titles do not provide sufficient information on this topic.

The two-volume set LNCS 11508 and 11509 constitutes the refereed proceedings of of the 18th International Conference on Artificial Intelligence and Soft Computing, ICAISC 2019, held in Zakopane, Poland, in June 2019. The 122 revised full papers presented were carefully reviewed and selected from 333 submissions. The papers included in the first volume are organized in the following five parts: neural networks and their applications; fuzzy systems and their applications; evolutionary algorithms and their applications; pattern classification; artificial intelligence in modeling and simulation. The papers included in the second volume are organized in the following five parts: computer vision, image and speech analysis; bioinformatics, biometrics, and medical applications; data mining; various problems of artificial intelligence; agent systems, robotics and control.

In recent years data streams became a ubiquitous source of information, and thus stream mining emerged as a new field in database research. Due to the inherently dynamic nature of data streams, stream mining algorithms benefit from being adaptive to changes in the properties of a data stream. In addition, when stream mining is done in a dynamic environment like a data stream management system or a sensor network, stream mining algorithms also profit from being adaptive to the changing conditions in this environment. This work investigates two kinds of adaptivity in data stream mining. First, a model for quality-driven resource adaptive stream mining is developed. The model is applied to stream mining algorithms so they efficiently utilize available resources to achieve mining results of the highest quality possible. Every stream mining algorithm is unique in its parameters, quality measures, and resource consumption patterns. We generalize these characteristics and develop a model that captures the interactions and correlations between variables involved in the stream mining process. We then express resource adaptive stream mining as a multiobjective optimization problem and use its solution to tune the input parameters of stream mining algorithms, which results in high quality mining and optimal resource utilization. The second topic investigated in this work is feature adaptive stream mining, which is concerned with adjusting the focus of the mining process to interesting features detected in the data stream. This research is motivated by the need to efficiently detect environmental phenomena from sensor data streams. We propose methods to detect and predict heterogeneous outlier regions, which represent areas of environmental phenomena of different intensities. With the help of predictions about the location and size of outlier regions, the sampling rate of individual sensors is adapted such that sensors in the vicinity of environmental phenomena obtain new measurements more frequently than other sensors in the network to allow for a precise and timely region tracking. The research in this work enhances the state-of-the-art in data stream mining as it makes stream mining algorithms more flexible to adapt to changes in the data stream and the mining environment.

Now in its second edition, this book focuses on practical algorithms for mining data from even the largest datasets.

The two-volume set CCIS 1142 and 1143 constitutes thoroughly refereed contributions presented at the 26th International Conference on Neural Information Processing, ICONIP 2019, held in Sydney, Australia, in December 2019. For ICONIP 2019 a total of 345 papers was carefully reviewed and selected for publication out of 645 submissions. The 168 papers included in this volume set were organized in topical sections as follows: adversarial networks and learning; convolutional neural networks; deep neural networks; embeddings and feature fusion; human centred computing; human centred computing and medicine; human centred computing for emotion; hybrid models; image processing by neural techniques; learning from incomplete data; model compression and optimization; neural network applications; neural network models; semantic and graph based approaches; social network computing; spiking neuron and related models; text computing using neural techniques; time-series and related models; and unsupervised neural models.

This book constitutes the revised selected papers of the 10th International Conference on Cloud Computing, Big Data & Emerging Topics, JCC-BD&ET 2022, held in La Plata, Argentina*, in June-July 2022. The 9 full papers were carefully reviewed and selected from a total of 23 submissions. The papers are organized in topical sections on: Parallel and Distributed Computing; Machine and Deep Learning; Cloud and High-Performance Computing, Machine and Deep Learning, and Virtual Reality.