Research And Design On Models Of Synthetic Evaluation Decision Support System

Management information system well deals with structural question. But in fact many half-structural and non-structural questions need to be solved. Under the circumstance, decision support system has been emerged. Synthetic decision support system is a new development direction with appearance of data warehouse, on-line analytical processing and data mining. Decision runs through the all course of management. In management process, choosing a good method among many solutions often appears. Only rational synthetic evaluation can produce correct choice. Synthetic evaluation is precondition of decision. No scientific synthetic evaluation no decision.The article outlines relative theories and technologies of synthetic evaluation decision support system and expatiates on model base. Then it deeply analyses and researches fuzzy math, grey theory and neural network. In allusion to request and character of instruction level, two-level fuzzy synthetic evaluation model, grey relation analysis model and neural network model are built. The article analyses model capability in detail and compares effects of two-level fuzzy evaluation model and neural network model to instruction level evaluation. Through example demonstration, these ways is effective to evaluate instruction level. Neural network model are more objective, just and fair to evaluate instruction level. It has theory significance and application value to instruction level evaluation. At last, the article builds system structure of synthetic decision support system and corresponding design of data warehouse, on-line analytical processing and data mining. This has direction significance to develop synthetic decision support system.