基于遗传算法和神经网络的软件界面美感建模
A new software interface aesthetics model based on genetic algorithm and neutral network

用户界面作为人与机器之间信息交互的桥梁, 良好的用户界面能够提高人机交互的效率. 本文以软件界面布局和色彩特征作为研究对象, 采用神经网络方法建立特征与软件美感之间的关系模型, 遗传算法进行特征选择, 找出对界面美感影响较大的特征, 并且通过主成分分析, 最终提炼出界面美感的设计要素.
A well designed software interface can increasingly improve the efficiency of the human and computer interaction(HCI) process. This paper focuses on the layout and color characteristics of interface. In order to discover the features which have the most influence on interface aesthetic feeling, we adopt machine learning method to build the relationship model between those characteristics and aesthetics of software interface, and then use the genetic algorithm(GA) to extract the important characteristics.At last, the aesthetic design elements of software interface can be refined by the method of the kernel principal component analysis(KPCA) applied to get the optimal feature subs