Support Vector Machine Based on Incremental Learning for Malware Detection

摘要: The training of traditional SVM method requires the solution quadratic programming, and consumes high memory has low speed for large data training. Incremental learning is one meaningful methods to continuously update learning, which keeps previous results, re only additional data, so as form a continuous process. This paper will study support vector machine based on incremental its application in malware detection. experiments carried out Internet Security Laboratory at Kingsoft Corporation suggested that, number virus samples, our can rapidly effectively sample features, avoids duplication history samples ensures prediction ability detection model. Keywords—Support Vector Machine; Learning; Malware Detection