A Neural Network-Based Agent Framework for Mail Server Management

Abstract

Amidst the era of e-economy, one of the difficulties from the standpoint of the information systems manager is, among others, the forecast of memory needs for the organization. In particular, the manager is often confronted with maintaining a certain threshold amount of memory for a prolonged period of time. However, this constraint requires more than technical and managerial resolutions, encompassing knowledge management for the group, eliciting tacit knowledge from the end users, and pattern and time series analyses of utilization for various applications. This paper proposes a framework for building an automated intelligent agent for memory management under the client-server architecture. The emphasis is on collecting the needs of the organization and acquiring the application usage patterns for each client involved in real time. Due to the dynamic nature of the tasks, incorporation of a neural network architecture with tacit knowledge base is suggested. Considerations for future work associated with technical matters comprising platform independence, portability, and modularity are discussed.

Data Dimension Reduction through Linearly Separable Patterns using Neural Networks

“…A Neural Network (NN) typically processes large-scale problems in terms of dimensionality, amount of data handled, and the volume of simulation or neural hardware processing [4]. It emerged as an area of Artificial Intelligence (AI) [5] to mimic the human neurons in both perception and learning.…”

“…Often, the inputs, outputs, and the computed gradient are deterministic, however, the minimization of error proceeds over all its random realizations. As a result, most supervised learning algorithms reduce to stochastic minimization of error in multi-dimensional weight space [4].…”

“…The weighting coefficients w 1 , w 2 , …, w n+1 of the dichotomizer can now be obtained by comparing equations (4) and (5) as follows:…”

“…Generation of linearly separable patterns or clusters is equally important for information systems as well as for conventional business entities such as project planning and/or manufacturing. By incorporating the suggested models of this paper into building "intelligent software agents" [4]. [9], a comprehensive Information Systems Management (ISM) may be achieved, allowing real-time network designs as well as frequent modifications to their topologies.…”

Intelligent System for Information Security Management: Architecture and Design

“…Artificial neural networks techniques are suggested for biometric identification applications (Kung, Mak & Lin, 2005). A system based on neural networks agent for mail server management is discussed in Willow (2005).…”