基于遗传算法和灰色关联分析的击键特征识别算法

基于用户击键特征的身份认证比传统的基于口令的身份认证方法有更高的安全性，现有研究方法中基于神经网络、数据挖掘等算法计算复杂度高，而基于特征向量、贝叶斯统计模型等算法识别精度较低。为了在提高识别精度的同时有效降低计算复杂度，在研究现有算法的基础上提出了一种基于遗传算法与灰色关联分析的击键特征识别算法。该算法利用遗传算法根据用户训练样本确定表征用户击键特征的标准特征序列，通过对当前用户击键特征序列与标准特征序列进行灰色关联分析实现用户身份认证。实验结果表明，该算法识别精度达到神经网络、支持向量机等算法的较高水平，错误拒绝率与错误接受率分别为0%与1.5%。且计算复杂度低，与基于特征向量的算法相近。

User authentication algorithm with keystroke features based on genetic algorithms and grey relational analysis

User authentication based on keystroke dynamics features is more secure than conventional user authentication approach only based on passwords.The neural network and data mining-based methods present high authentication accuracy,but have a high computational cost.The statistical and vector-based methods have shown low computational complexity,but are less accurate in user authentication.In order to improve authentication accuracy and reduce computational complexity synchronously,a new user authentication approach based on keystroke patterns was proposed.In the proposed approach,Genetic algorithm was employed to generate the common keystroke pattern of each user from the training set consisting of the user's normal keystroke samples.Then Grey Relational analysis method was applied to calculate the degree of grey slope incidence between common keystroke pattern and current keystroke pattern,the resultant value was compared with a threshold value determined by experiment to implement user authentication.Experimental results show this approach represents the same user authentication accuracy as neural network and data mining-based methods in terms of False Acceptance Rate(FAR) and False Rejection Rate(FRR),false acceptance rate and false rejection rate of this method are 1.5% and 0% respectively.It is also shows that the computational complexity of the proposed method is lower than that of some other methods.