观察式学习条件下类别维度与家族相似性类别学习

在观察式学习条件下考察类别维度数量（三维度类别、六维度类别、九维度类别），类别维度结构（线性分离结构、非线性分离结构）的变化对家族相似性类别学习的影响。结果表明：观察式学习条件下，类别维度数量的变化对样例学习无影响，对特征学习有影响，表现在特征正确再认数量上有差异，但特征再认正确率上无差异；类别维度结构的变化对样例学习有影响，对特征学习无影响，表现在样例正确再认数目上线性结构大于非线性结构。

CATEGORY DIMENSIONALITY AND CATEGORY LEARNING OF FAMILY RESEMBLANCES CATEGORY ON THE CONDITION OF OBSERVING LEARNING

On the condition of observing learning, the authors examined the effect on category learning of family resemblances which the change of numbers and structures of category dimensionality have. The results suggest that there was no significant differences between exemplar learning of more category dimensionality and less category dimensionality on the condition of observing learning, but there was significant differences for numbers of features′ recognizing between more category dimensionality and less category dimensionality. The exemplar learning of linearly separable structure was significantly defferent from no-linearly separable structure, but the differences for number of features′ recognizing were not significant between linearly separable structure and no-linearly separable structure.