基于支持向量机的光刻胶粘接芯片存储安全检测

Storage Security Detection of Photoresist Adhesive Chips Based on Support Vector Machine

光刻胶因其良好的性能被运用于芯片中,是芯片制作不可缺少的重要材料。对于保证光刻胶粘接芯片的存储安全,需要对其进行检测。文章将基于支持向量机检测光刻胶粘接芯片存储安全,为了提高检测的准确率,引入深度置信网络。通过仿真实验的方法,研究支持向量算法、深度置信网络算法、两者相结合的算法对光刻胶粘接芯片存储安全进行检测。研究结果表明支持向量算法的检测准确率低于深度置信网络算法低于两者相结合的算法,即两者相结合的算法检测光刻胶粘接芯片存储安全的准确率更高、误报率更低。

Photoresist is used in chips because of its good performance,and is an indispensable material for chip production. To ensure the storage security of the photoresist bonding chip,it needs to be tested. In this paper,the support vector machine is used to detect the storage security of the photoresist bond chip. In order to improve the accuracy of the detection,a deep belief nets is introduced.Through the method of simulation experiment,the support vector algorithm,deep belief nets algorithm and the combination of the two are used to detect the storage security of the photoresist bonding chip. The results show that the detection accuracy of support vector algorithm is lower than that of deep belief nets algorithm,and that of deep belief nets algorithm is lower than that of the combination of the two algorithms. That is,the combination of the two algorithms has higher accuracy and lower false alarm rate in detecting the storage security of bonded chips.