2,923 Hits in 3.8 sec

Concept Drift and Anomaly Detection in Graph Streams

Daniele Zambon, Cesare Alippi, Lorenzo Livi
2018 IEEE Transactions on Neural Networks and Learning Systems  
Here, we consider stochastic processes generating graphs and propose a methodology for detecting changes in stationarity of such processes.  ...  In addition, we provide a specific implementation of the methodology and evaluate its effectiveness on several detection problems involving attributed graphs representing biological molecules and drawings  ...  Concept Drift and Anomaly Detection in Graph Streams Daniele Zambon, Student Member, IEEE, Cesare Alippi, Fellow, IEEE, and Lorenzo Livi, Member, IEEE Abstract-Graph representations offer powerful and  ... 
doi:10.1109/tnnls.2018.2804443 pmid:29994077 fatcat:hbulbnosajgptfemcegtenwomi

The Application of a Double CUSUM Algorithm in Industrial Data Stream Anomaly Detection

Guang Li, Jie Wang, Jing Liang, Caitong Yue
2018 Symmetry  
Compared with automatic outlier detection for data streams (A-ODDS) and with sliding nest window chart anomaly detection based on data streams (SNWCAD-DS), the DCUSUM-DS can account for concept drift and  ...  This paper proposes a data stream anomaly detection algorithm combined with control chart and sliding window methods.  ...  The difficulty of data stream anomaly detection lies in concept drift. The future distribution of a data stream is unknown.  ... 
doi:10.3390/sym10070264 fatcat:4xavlltw2jdhzdmjcbdnlgtwnq

Application of Sliding Nest Window Control Chart in Data Stream Anomaly Detection

Guang Li, Jie Wang, Jing Liang, Caitong Yue
2018 Symmetry  
detection based on the data stream (SNWCAD-DS) by employing the concept of the sliding window and control chart.  ...  ratio and classifies the conceptual drift data stream online.  ...  The data stream anomaly detection includes concept drift [16] and drift level [17] detection, which is an online classification problem.  ... 
doi:10.3390/sym10040113 fatcat:e2t2ygn6dzhjjiqjirwfrv54ma

Multi Novel Class Classification of Feature Evolving Data Streams with J48

Punam D., A.M. Dixit
2015 International Journal of Computer Applications  
In the Data stream classification main issues are infinite length, concept drift, concept development, and feature development.  ...  In the existing system of data stream method researcher tackle on the only two issues i.e. concept drift and concept evolution problem of classification.  ...  Initially in [13] studied the novel class detection problem in the neighbourhood of concept-drift and infinite length.  ... 
doi:10.5120/ijca2015905652 fatcat:bgxakksgfjdlfcgdyf7eahodky

Introduction to the special issue on discovery science

Michelangelo Ceci, Toon Calders
2018 Machine Learning  
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.  ...  The paper introduces the TORNADA framework, in which multiple diverse classifiers and drift detection algorithms are executed in parallel.  ...  Stacking Fast Hoeffding Drift Detection Methods for Evolving Data Streams" by Pesaranghader et al. focuses on adaptive learning from evolving data streams.  ... 
doi:10.1007/s10994-018-5749-6 fatcat:aisy73tcirexxnd45d5b34wyy4

An Intrusion Detection System for the Internet of Things Based on Machine Learning: Review and Challenges

Ahmed Adnan, Abdullah Muhammed, Abdul Azim Abd Ghani, Azizol Abdullah, Fahrul Hakim
2021 Symmetry  
We present the three main challenges of an IDS, in general, and of an IDS for the Internet of Things (IoT), in particular, namely concept drift, high dimensionality, and computational complexity.  ...  This article concludes that three elements of concept drift, high-dimensional awareness, and computational awareness that are symmetric in their effect and need to be addressed in the neural network (NN  ...  In fuzzy-based anomaly detection, the rule-based is connected to inputs.  ... 
doi:10.3390/sym13061011 fatcat:zpq3ds2ifnfmxddbnhxuej7knm


2013 International journal on artificial intelligence tools  
This paper applies ensemble-based stream mining, supervised and unsupervised learning, and graph-based anomaly detection to the problem of insider threat detection.  ...  Ensemble-based stream mining leverages multiple classification models to achieve highly accurate anomaly detection in such streams, even when the stream is unbounded, evolving, and unlabeled.  ...  data as a graph and apply unsupervised learning to detect anomalies.  ... 
doi:10.1142/s0218213013600130 fatcat:qntmuz7zrjaanl56qfaajvcrra

Anomaly, event, and fraud detection in large network datasets

Leman Akoglu, Christos Faloutsos
2013 Proceedings of the sixth ACM international conference on Web search and data mining - WSDM '13  
Detecting anomalies and events in data is a vital task, with numerous applications in security, finance, health care, law enforcement, and many others.  ...  The goal of this tutorial is to provide a general, comprehensive overview of the state-of-the-art methods for anomaly, event, and fraud detection in data represented as graphs.  ...  That is, methods in this topic are grouped into outlier, anomaly, fraud, event, change, drift, fault detection separately.  ... 
doi:10.1145/2433396.2433496 dblp:conf/wsdm/AkogluF13 fatcat:i7m6c3g7j5agvnova3javx3aly

Guest Editorial: Non-IID Outlier Detection in Complex Contexts

Guansong Pang, Fabrizio Angiulli, Mihai Cucuringu, Huan Liu
2021 IEEE Intelligent Systems  
We believe this work provides an example of joint classification and outlier detection on imbalanced streaming data with potential concept drift.  ...  In "Anomaly detection aided budget online classification for imbalanced data streams," 8 Liang et al. study the problem of classification on imbalanced streaming data in the presence of outliers.  ... 
doi:10.1109/mis.2021.3072704 fatcat:z7ohc4yue5hw7gmsauflfwklqu

Ensemble-Based Online Machine Learning Algorithms for Network Intrusion Detection Systems Using Streaming Data

Nathan Martindale, Muhammad Ismail, Douglas A. Talbert
2020 Information  
complexity, and response to concept drifts.  ...  drift in the most effective way.  ...  between run-time performance, anomaly-detection accuracy, and response to concept drifts.  ... 
doi:10.3390/info11060315 fatcat:ly5wmmmxpzbb7ltjp3mn65u6sq

Analyzing Performance of Classification Algorithms on Concept Drifted Data Streams

Aradhana Nyati, Divya Bhatnagar, Avinash Panwar
2017 International Journal of Computer Applications  
In this paper the main issue of concept drift is addressed with real and synthetic data streams and the comparison of ensemble classifiers has been made in view of concept drift for the assessment of the  ...  Various classifiers were applied on data stream with and without concept drift for analysis.  ...  Sudden drift in the graph shows the occurrence of concept drift in the data streams.  ... 
doi:10.5120/ijca2017913065 fatcat:u7yda2myobcxdjvgkxxfztcqb4

Data stream mining techniques: a review

Eiman Alothali, Hany Alashwal, Saad Harous
2019 TELKOMNIKA (Telecommunication Computing Electronics and Control)  
In this paper, we review real time clustering and classification mining techniques for data stream.  ...  Analyzing this massive data in real-time and extracting valuable knowledge using different mining applications platforms have been an area for research and industry as well.  ...  KME [33] is a recent classifier leverages supervised and unsupervised knowledge to detect concept drift and recognizes recurrent concepts.  ... 
doi:10.12928/telkomnika.v17i2.11752 fatcat:rls2qzcl3vhobmkpycsdwhzplu

Analysis of Classification and Clustering based Novel Class Detection Techniques for Stream Data Mining

Kamini Tandel, Jignasa N. Patel
2015 International Journal of Engineering Research and  
Concept drift means data changes rapidly over time and novel class define as new class appear in continuous data stream.  ...  Due to its dynamic changing nature it has some major challenges like infinite length, novel class detection and concept-drift.  ...  Concept drift that means data always changes over time and concept evolution means new class arrived in data stream.  ... 
doi:10.17577/ijertv4is100160 fatcat:cqtpjn4qxrb4pndtteg4uhxlfu

Concept Drift Detection Based on Anomaly Analysis [chapter]

Anjin Liu, Guangquan Zhang, Jie Lu
2014 Lecture Notes in Computer Science  
In this paper, we propose a novel concept drift detection method, which is called Anomaly Analysis Drift Detection (AADD), to improve the performance of machine learning algorithms under non-stationary  ...  Experiments illustrate that this AADD method can detect new concept quickly and learn extensional drift incrementally.  ...  Secondly, in Section 4.2, we show the accuracy change caused by concept drift and plot the anomaly points at each time step on a graph.  ... 
doi:10.1007/978-3-319-12637-1_33 fatcat:4d2zjzl5ordzzaxui5eka27ywe

Identifying Concept-drift in Twitter Streams

C.S. Lifna, M. Vijayalakshmi
2015 Procedia Computer Science  
Identification and handling of this concept-drift in such Big Data Streams is present area of interest.  ...  This study puts forward a novel approach towards identifying concept-drift by initially grouping topics into classes and assigning weight-age for each class, using sliding window processing model upon  ...  In short, Concept-Drift Analysis is the integrated study of identifying and handling Concept-Drift in this evolving stream of data.  ... 
doi:10.1016/j.procs.2015.03.093 fatcat:bzaaw2z6tnhtzn4qh3sjlqmvii
« Previous Showing results 1 — 15 out of 2,923 results