|
|
Applied Technology
BPA (Business Process Automation)
• Artificial Intelligence
- Expert Systems
- Inference Engine
- Business Rules
- Complex Scenarios
- Neural Networks
- Prediction
- Cause – Effect
- Clustering
- “What-If” Scenarios
• Business Intelligence
- EIS (Executive Information Systems)
- OLAP (On-Line Analytical Processing)
- Data Warehouse
• Componentization
• WEB – Java
• Object Orientation
INFORMATION...
i - ARTIFICIAL INTELLIGENCE - AI objective is to empower computers to be able to reason in a similar way a human being would. There are different branches in the AI context. One of them is the Expert Systems or Rule Based Systems and another one is the Neural Networks.
i.1 - EXPERT SYSTEMS or RULE BASED SYSTEMS - They are applications designed to solve problems that are solved by human experts in a particular field of knowledge. To solve such problems Expert Systems need to access a knowledge base that contains the problem domain. This knowledge is created using Expert Systems development tools in the most efficient manner. Expert Systems also need to explore one or more reason mechanisms in order to apply its knowledge to solve such problems. This reason mechanism is implemented in Expert Systems development tools in the form of an algorithm called INFERENCE ENGINE that is capable of inferring knowledge that is not explicitly stored in the knowledge base. One of the most important features of an Expert System is the explanation capability. In the same way a human expert can explain his line of reason to reach some conclusion, so can an Expert System explain the sequence of rules the INFERENCE ENGINE applied to reach its objective.
INFERENCE ENGINE – Algorithm that, by itself, organizes and sequences the firing of business rules that are inside the knowledge base. To accomplish this the IE uses different techniques, such as forward chaining, backward chaining and pattern matching. IE technology is always present in Expert Systems Development Tools.
i.2 – NEURAL NETWORK BASED SYSTEMS – They are applications inspired in brain structures, capable of processing information. Neural Network models are the answer to problems that deal with common sense reason, fuzzy logic and incomplete or inconsistent information processing.
The main characteristics of Neural Network based Systems are:
Ability to learn from examples and to generalize this learning, in a way it is able to recognize similar examples that were never presented to the Neural Network before;
Good performance in poorly defined tasks, where explicit knowledge in how to find a solution is missing;
Does not require mathematical models knowledge nor application domain rules knowledge from the user;
High noise immunity, meaning a Neural Network does not collapse in the presence of fake or missing information. It only degrades its performance in a gradual manner;
It puts stored data into work, transforming them into useful business information.
i-3 - INTELLIGENT TOOLS – Software Tools that enable the development of Expert Systems and Neural Network based Systems.
ii - PARAMETRICAL SYSTEMS – They are flexible applications that allow the end-user, by using friendly screens, to modify application parameters in order to rapidly adequate the application to new business conditions.
iii - INTELLIGENT SOLUTIONS – They are flexible applications that combine Expert Systems with Neural Networks and Parametrical Systems. They are today important players in the fast business transformation environment, enabling powerful business process automation. They can process and automate business instead of just registering business facts. They can predict future business indicators and take actions before the problem take place.
Contact us...
|
|
|
|
|
|
|
