1. Most interpolative reasoning methods in multi-dimension cannot guarantee the convexity and normality of result, and cannot consider weights of each variable influencing conclusion.

2. Many different interpolative reasoning approaches have been proposed to one dimension condition through interpolative reasoning theory in recent years.

3. Aim Fuzzy interpolative reasoning has become an important approach to solve the problem of sparse rules.

4. The improved algorithm of optimal interpolative estimation was adopted under the condition of strong signal interrelation.

5. Furthermore, by altering pace to select interpolative dots, this method can avoid selecting evil dots to advance the effectiveness.

6. Methods An interpolative reasoning method by considering the location and the shape of fuzzy sets of all premises and consequences of the known rules at the same time was presented.

7. The aim of this study was to establish a analytical method for furosine that determine whether reconstituted milk was interpolative in UHT milk, and measure furosine content in raw milk and UHT milk.

8. Kriging is an important part of geostatistics, which deals with spatially distributed data, the estimation of Kriging is linear optimal and unbiased interpolative.

9. It is also difficult to keep convexity and normality using KH linear interpolative reasoning method.

10. They are with many good experience, but the research for interpolative reasoning under multidimensional sparse rules condition is lacking and a few existing approaches have some faults.

11. In order to get better result when rule base is sparse, we propose a similarity interpolative reasoning method which can keep the convexity and normality of the reasoning result better.