This preview shows page 1. Sign up to view the full content.
Unformatted text preview: transactions. Neural network:- yes, definitely suitable. Neural networks are programmed to learn from past history, and so would be very good at examining past patterns of transactions by customer to determine whether there are any transactions that appear to be fraudulent. Fuzzy logic:- no, not suitable. These systems simulate human reasoning, and are designed to have computers behave less precisely. To identify potential fraud transactions, the computer systems need to be specific in identifying the potentially fraudulent transactions. Other suitables:- most types of information systems that we have discussed could be suitable, for example an MIS system could provide exception reports, a DSS could use statistical analysis to identify potential at risk transactions...
View Full Document
This note was uploaded on 04/11/2011 for the course ADMS 2511 taught by Professor Jiu during the Spring '09 term at York University.
- Spring '09