Forecast system of pear scab management based on case-based reasoning and fuzzy ISODATA clustering

Pear scab caused by Venturia nashicola Tanaka et Yamamoto is one of the main diseases in the Centre and North China and sometimes do a severe damage in pear production there. A forecast system of pear scab management was established on the basis of Case-based Reasoning and Fuzzy ISODATA Clustering. Every typical case was recorded in the database according to the incident status of pear scab and the synchronic meteorological data in area of the Ancient Yellow River Valley and expressed as a set of feature collection, which was composed of the date, rainfall, air temperature, atmospheric humidity, the percentage of diseased leaves and fruits. Fuzzy ISODATA clustering was utilized in the reasoning to imitate the human expert better. The System was realized by programming using the software Visual Basic version 6.0. The practical application showed that the forecast system had a high predictability and accuracy and could provide a useful reference to the pear growers in control of pear scab.