DataX™ SPECIFICATIONS
- An Intelligent Data Mining Software for Business & Engineering

Introduction to DataX™
Based on neural network, fuzzy logic, non-linear programming, and time series analysis methodologies, DataX(tm) is a complete software suite, ideal for business planning, financial forecasting, customer pattern matching, market trend analysis, performance evaluation, case study  simulations, optimization, and resources scheduling. It can also be used in process controls.

Feature Highlights
Object-oriented component design configurable for custom applications
Interactive system modeling tools with multimedia capability (text, sound, pictures)
Natural language programming environment
GIS-embedded with map interface
Hybrid AI technologies embedded (neural-fuzzy, time series analysis etc)
Accessible to major databases using ODBC or by DBF conversion
Applicable to business management, marketing, finance, and engineering fields

Data Analysis Tool Base
Fuzzy GMDH and regression analysis
Fuzzy AHP and multivariate analysis
iDSS (intelligent decision support system) and fuzzy expert systems
Trend analysis, cyclic component extraction and forecasting
Fuzzy correlation matrix, clustering and categorization
Automated model building by neuro-fuzzy based system identification
Fuzzy utility and demand functions
Linear and non-linear programming
Optimization and scheduling using genetic algorithms

Modular sub-systems configurable for custom designs
Model base
Knowledge base
Query, sorting and search engines
Analysis tools base
Data Import/Export (ODBC)

Object-oriented Software Architecture: Objects Hierarchy
Entity objects
Relation objects
Problem objects
Operator objects
Optimization  objects

User-Defined Graphic Desks for Different Levels of an Organization
Executive Desk
Department manager Desk
System Analysis Desk
Computer Administrator's Desk

ODBC to major databases
Oracle, Sybase, Informix, FOxPro, SQL, dBase II, IV and others.

Data visualization in color
Bar charts
Plots
Tables, labels
Spreadsheets
HTML pages
Java applets
Text files
Digital images (Bitmap, GIF, JPEG)

Result reporting and publication
Contents and format configurable by system admin
Pushed to various levels of an organization (as different GUIs)
HTML and Java applets publication capability

Diverse Applications
Goal determination and alternative evaluation via simulations
System modeling
Analysis and evaluation
Optimization and decision support systems
Controls (behaviors, investment, product quality, etc.)
Customer consumption preference analysis using
Consumer spending pattern matching

Example-1: Commodity price forecasting (below), the real-world sample.
Upper-left - original data for 5 commodity prices for 5 years (5 data from from history)
Upper-right - principle component analysis (3 major harmonics each of the 5 data)
Lower-left - build a model for each data and reconstruct data with the models (green), error < 3%
Lower-right - forecasting (black) using the models vs. real-world prices (color) of 5 goods.

Example-2: Neuro-fuzzy investment performance simulation study - a real case (below)
Explanation - this example illustrates how to DataX(tm)'s neuro-fuzzy tools to perform system modeling, forecasting and simulations (3 inputs factors on the left-hand side, system model in the middle, and 3 investment performance indices on the right-hand side). This is an ideal to tool to model alternatives or various scenarios of different business applications, such as investment return prediction, market fluctuation simulation, stock or mutual fund analysis, consumption categorization, pattern matching, etc.

 
© 1997-2001, ZAPTRON Systems, Inc.