Objectives: Introduction Over View of System Analysis and Design
Download 0.94 Mb. View original pdf
|
- Navigate this page:
- 6.4.4.1 Avoiding problems with Decision Trees
| 6.4.4 Identifying Data Requirements We have already pointed out the use of decision trees to formally highlight the sequential nature of many business decision, and we have shown that decision trees are effective when describing business problems of more than one dimension or condition. However, they also identify critical data requirements surrounding the decision process that is, they indicate the sets of data the manager requires to formulate decision or select action. The explicit data in the payment example are payment data, amount of invoice, and discount allowance percentage. There are other important data elements such as invoice details (number, supplier name and address, new invoice amount payable, and adjustments to discount taken that are indict (not directly expressed in the decision tree. The analyst must identify and list all the data used in the decision process, even must identify and list all the data used in the decision process, even though the decision tree does not show all the individual data items. It decision trees are assembled after the completion of data flow analysis (which tracks the flow of data through busbies processes, critical data may already be defined in the data dictionary (which describes system data and where they are used. If decision trees are identify each data element needed to make a decision trees are identify each data element needed to make a decision. The data dictionary format, is useful for listing and describing data elements as they are identified and understood. The date requirements discussed for decision trees also apply to the other decision – analysis methods that will be discussed. Analysis’s need to describe and define all data used indecision making, so that the system can be designed to produce data properly. 6.4.4.1 Avoiding problems with Decision Trees Decision trees may not always be the best tolls for decision analysis. A decision tree fora complex system with many sequences of steps and combinations of conditions will be unwieldy. A large number of branches with many paths through them will could rather than aid analysis. The analyst may not be able to determine which business policies or practices guide the formulation of specific decisions. Where these problems arise, decision tables should be considered. Download 0.94 Mb. Share with your friends:
|