Volume1, Number2, 2000


 

 

Title:

Neuro-Fuzzy Networks: Adaptive Fuzzy Modeling and Control

Author:

Tang Nan, Fei-Yue Wang, Frank W.Ciarallo, and Guihe Qin

Abstract:

NFNs are knowledge-based multi-layer neural networks constructed by integrating three types of modular sub-nets for pattern recognition, fuzzy reasoning, and control synthesis, respectively. In this way, a NFN combines the reasoning procedure of fuzzy logic and learning capability of neural networks uniquely, and thus is able to incorporate linguistic knowledge in the form of fuzzy rules in its network structure and then refine this knowledge through training and self learning of the networks. The application of NFN for modeling and control of unknown nonlinear dynamic processes is discussed here with simulation results presented to illustrate these ideas.

 

 

Title:

An Object Oriented Approach to Multidimensional Databases & OLAP Operations

Author:

Juan Trujillo, Manuel Palomar, and Jaime Gómez

Abstract:

In this paper, we extend our first proposal called Object Oriented Multidimensional Data Module (OOMD) presented by J. Trujillo and M. Palomar (1998) in [9] providing On-line Analytical Processing (OLAP) operations to allow a subsequent data analysis. The basic elements in our model are dimension classes and fact classes to represent the basic objects in a Multidimensional Database (MDB). Basic cubes (cube classes) are then built from these previous classes, which allow us to carry out the subsequent data analysis. We then apply OLAP operations to these basic cubes to define views that will enable us to analyze data multidimensionally. We also provide a design methodology to define all possible OLAP operations to be applied to cube class objects, which provides the user a more restrictive way of analyzing data. On the one hand, our OOMD provides a view of the whole database scheme as well as considering specific cubes for user particular analysis. On the other, our approach is based on the OO Paradigm, which provides classes to encapsulate both static and dynamic properties of data (OLAP operations).

 

 

Title:

Using Neural Networks To Improve a Market Timing Heuristic

Author:

William Leigh, Ross Hightower, and Ron Rubin

Abstract:

We implement a pattern-based heuristic from stock market technical analysis and perform back-testing with price and volume time series data for publicly listed stocks. A subsequent application of back-propagation neural networks improves the trading results by reducing the probability of making purchases when the stock price is to go down. The research investigates the validity of stock market technical analysis, which is generally considered to be inconsistent with the efficient market theory, and explores the combination of neural network techniques with conventional pattern recognition for the implementation of expert system heuristic rules.

 

 

Title:

Software Architecture Analysis: A Dynamic Slicing Approach

Author:

Taeho Kim, Yeong-Tae Song, Lawrence Chung, and Dung T. Huynh

Abstract:

As the complexity of software systems increases, so does the need for a good mechanism of abstraction. Software architecture design is an abstraction, hiding an immense amount of details about the data structures, algorithms, idiosyncrasies of programming language constructs, etc. that may be used in implementing the system-to-be. Fundamental as it may be to the modeling of the system, the very nature of this high level abstraction can also pose difficulties with the understanding and analysis of the behavior of the system-to-be. This paper introduces the notion of dynamic software architecture slicing (DSAS) in order to alleviate such difficulties. A dynamic software architecture slice represents the run-time behavior of those parts of the software architecture that are selected according to a particular slicing criterion such as a set of resources and events. This paper also describes a methodology for using the notion, and an algorithm to generate dynamic software architecture slices. The feasibility and the expected benefits of the approach is demonstrated through a study of part of an electronic commerce system and a run-time execution of its architecture using a tool.