[1]易茂祥,宋晨钰,于金星,等.基于随机森林的集成电路适应性测试方研究[J].郑州大学学报(工学版),2021,42(04):13-18.
 Adaptive Test Method ba<x>sed on Random Forest Algorithm[J].Journal of Zhengzhou University (Engineering Science),2021,42(04):13-18.
点击复制

基于随机森林的集成电路适应性测试方研究()
分享到:

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
42
期数:
2021年04期
页码:
13-18
栏目:
出版日期:
2021-07-30

文章信息/Info

Title:
Adaptive Test Method ba<x>sed on Random Forest Algorithm
作者:
易茂祥宋晨钰于金星宋钛鲁迎春黄正峰
文献标志码:
A
摘要:
随着集成电路产业的高速发展,芯片的复杂度不断增加,不可避免地导致高测试成本问题。适应性测试为降低测试成本提供了一种解决方案。提出了一种基于随机森林的适应性测试方法,利用特征重要性筛选出最为重要的测试组,同时根据测试组所能测出缺陷芯片的个数对测试组进行排序,通过删除部分测试组来降低测试时间。实验结果表明,随机森林与KNN和逻辑回归相比始终保持较低的测试逃逸水平。与现有测试方法相比,测试组删除可以大幅降低测试时间,从而降低测试成本。
Abstract:
With the rapid development of the integrated circuit industry, the complexity of chips continues to increase, which inevitably leads to high test cost problems. Adaptive testing provides a solution for reducing testing costs. An adaptive test method ba<x>sed on random forest is proposed,which uses feature importance to screen out the most important test groups. At the same time, the test groups are ranked according to the number of defective chips that can be measured by the test group to achieve the purpose of reducing test time. Experimental results show that compared with KNN and logistic regression, random forest always keeps a low level of test escape. Compared with existing test methods, test group deletion can greatly reduce test time, thereby reducing test costs
更新日期/Last Update: 2021-08-26