Classification algorithms
Explorations in parallel distributed processing: a handbook of models, programs, and exercises
Explorations in parallel distributed processing: a handbook of models, programs, and exercises
Computer systems that learn: classification and prediction methods from statistics, neural nets, machine learning, and expert systems
Instance-Based Learning Algorithms
Machine Learning
C4.5: programs for machine learning
C4.5: programs for machine learning
IEEE Expert: Intelligent Systems and Their Applications
Knowledge discovery preprocessing: determining record usability
ACM-SE 36 Proceedings of the 36th annual Southeast regional conference
ANSWER: network monitoring using object-oriented rules
AAAI '98/IAAI '98 Proceedings of the fifteenth national/tenth conference on Artificial intelligence/Innovative applications of artificial intelligence
Knowledge discovery in databases: tools and techniques
Crossroads - Special issue on networks and distributed systems
Discovery of frequent DATALOG patterns
Data Mining and Knowledge Discovery
A Survey of Temporal Knowledge Discovery Paradigms and Methods
IEEE Transactions on Knowledge and Data Engineering
Investigating and Evaluating Behavioural Profiling and Intrusion Detection Using Data Mining
MMM-ACNS '01 Proceedings of the International Workshop on Information Assurance in Computer Networks: Methods, Models, and Architectures for Network Security
Industry: predicting telecommunication equipment failures from sequences of network alarms
Handbook of data mining and knowledge discovery
Statistical analysis and prioritisation of alarms in mobile networks
International Journal of Business Intelligence and Data Mining
Managing power conservation in wireless networks
ADMA'10 Proceedings of the 6th international conference on Advanced data mining and applications - Volume Part II
Behavioural Proximity Discovery: an adaptive approach for root cause analysis
International Journal of Business Intelligence and Data Mining
International Journal of Applied Metaheuristic Computing
Application of Bayesian Networks for Autonomic Network Management
Journal of Network and Systems Management
| Hi-index | 0.00 |
Modern business enterprises depend on computer-processed transactions. As the computer's role becomes more pervasive, businesses are attempting to warehouse huge volumes of historical data with the expectation of mining it for knowledge. These businesses hope to determine trends and patterns that could improve their organization's effectiveness, efficiency, and prospects. A large-scale telecommunication network, such as that of AT&T, processes many transactions that vary greatly over time. In this article, we examine how to warehouse data about faulty network behavior and how to later mine it to find trends and patterns that characterize current and future network behavior.Modern communication networks' increasing size and complexity will require intelligent systems to help manage and maintain them. For optimum efficiency, these systems should analyze and resolve problems in the network automatically. In addition, they should identify potentially serious problems before such problems degrade network performance. In this way, intelligent systems can greatly improve the network's reliability and quality. However, this can be extremely difficult. For example, AT&T's worldwide network is a massively interconnected structure of several complex devices with over a million different paths or circuits. During any day, millions of transmissions may occur over this network. Many sophisticated computer systems and devices interact to achieve extremely high reliability.There are various aspects of managing and maintaining such a network, including provisioning various services, managing resources and traffic, proactively maintaining the network, and troubleshooting different kinds of problems. Artificial intelligence techniques have proven useful in building systems that automate network management and maintenance. Such systems have automated the diagnosis and repair of network transmission problems. These earlier methods concentrated on incorporating knowledge obtained from domain experts in rule-based systems to troubleshoot problems. Such systems focused on resolving hard failures that had already occurred. However, there is an increasing demand for maintaining the network proactively. This includes the capability to predict problems that are likely to persist as well as those that are likely to degrade network performance. This proactive approach is very important for maintaining a high level of network quality and reliability and can provide a significant edge in the telecommunications service industry.Telecommunication system managers can practice proactive network maintenance by monitoring the performance of the network continuously over time and identifying patterns that indicate future problems. Monitoring network performance involves analyzing extremely large amounts of diagnostic data that varies with time. Only computer analysis will enable us to successfully look for behavioral patterns in such large volumes of data.In this article, we present an approach to systematically and exhaustively search time-varying, diagnostic data using machine-learning and correlation techniques. The correlation technique uses an AT&T system called Scout. We describe the application domain, explain the approach that we used to enable the machine to learn from time-dependent problems in the network, apply alternative methods of machine learning to our problem, and report our results. We illustrate the approach by specifically applying it to diagnostic data from transmission elements in AT&T's worldwide digital communication network. Nevertheless, although this article focuses on telecommunication networks having many homogeneous network components, our work can apply to other types of networks as well.