Knowledge Reductionand Rule Acquisition In Incomplete Interval-valued Information System

As the complexity of reality and the restricted cognition of people, some valuecouldn’t be decribed by certain number. In this case, we can only obtain a domain ofthis number, which is equal to the upper limit and lower limitof the number. As a result,interval value is introduced to describe this matter. On the other hand, it is difficult toget a set of complete information which means some null value is contained in thatinformation. So it matters a lot to research on incomplete interval-valued informationsystem as it is common in real life.In this paper, rough set theory was employed to process attribute reduction and ruleacquisition in incomplete interval-valued information system. Rough set theory was firstproposed by Z.Pawlak, it is a powerful math theory which is good at dealing withuncertain knowledge.Attribute reduction and rule acquisition are two important areas of rough set theory.There are two key points of this paper. One is to classify the objects in incompleteinterval valued information system.In this part, two methods were established based onsimilary relation and cluster analysis. The other is to set up the general system theory ofattribute reduction and rule acquisition in incomplete interval-valued informationsystem. Then the uncertainty measurement of this type of system was also established.A great deal of work has been done on incomplete information system andinterval-valued information system while less is done when the two kinds of systemcombined together. It is of great practicability to research on incomplete informationsystem and interval-valued information system. This paper provides a suggestive workfor further researching.