牛顿插值与分块算法对自适应支持度的研究

Research of Newton interpolation and block algorithm about adaptive support

为解决高时间复杂度和牛顿插值算法中的凸点问题,提出了一种分块算法、牛顿插值算法、Apriori算法及哈希表冲突处理方法思想等相互结合的方案。充分利用分块算法的分块特性,解决了高时间复杂度问题;通过利用哈希表冲突处理方法思想对点值误差进行修正,解决有关凸点问题。结合电子商务应用系统实例,对设计方案进行了研究、对比与验证,得到了一个低时间复杂度、算法运行稳定的可行方案,使关联规则挖掘结果的精度得到进一步的提高,从而增强了用户的决策和收益能力。

To resolve problems of high time complexity and bump, a solution of combining with block algorithm, Newton inter polation algorithm, Apriori algorithm and thought about conflict approach in hash table is proposed. The program takes full ad- vantage of the sub block of the block algorithm to solve the high complexity problem, and makes the good use of a approach to deal with the hash table＇s conflict to resolve the bump. Through research, comparison and verification of the design of the pro- gram applied in an e-commerce application, a viable solution is acquired, which has lower time complexity, the stable running al gorithm and higher accuracy, enhancing users＇ decision-making and earning power.