Virginia Integrated Pest Management Expert for Wheat
MetadataShow full item record
This is a rule-based system developed on the Microsoft Windows platform. Knowledge about crops and pest management is represented in the form of IF/THEN rules, demons, and "when-changed" methods. The inference engine analyzes specific crop system information entered by the user to determine potential risks of outbreak for wheat crop pests common to Virginia. These potential outbreak risks are presented as low, medium, and high levels of risk and are presented for each of 15 pests of wheat in Virginia.
The system was evaluated using thirty random cropping system scenarios. By comparing expert system output with output from human experts, it was shown that the expert system agreed with human expert opinions in 84 percent of the decisions made. Statistical analysis of the insect pest data showed that there was no significant statistical difference between the distribution of the human expert predictions and the expert system predictions. Statistical analysis of the disease pest data showed that there were some significant statistical differences between the distribution of the human expert predictions and the expert system predictions.
- Masters Theses