Data mining method based on computer forensics-based ID3 algorithm
Iu Qin · 2010
Data mining method based on computer forensics-based ID3 algorithm is presented in the study. Forensics data are unconstant, noisy and dispersive. Based on these characteristic of forensics data, the improved ID3 algorithm from adopting weight and two times information is gained. The examples can be used as the experiment data, and 100 test samples which is independent of training samples are applied to judge error rate of decision tree rules. The experimental results show that the error rate of ID3 is 8.9% and the error rate of improved algorithm is 5.4%, which indicates the accuracy of the proposed method is higher than ID3 algorithm. It can be seen that the improved method used in the computer forensics process is entirely feasible.