基于SVM的建筑学专业学生综合设计潜力评价  

An application of SVM neural network to evaluate comprehensive design potential of architecture graduates

作　　者：吴蔚[1] 吴农[2] Wu Wei;Wu Nong(School of Architecture and Urban Planning,Nanjing University,Nanjing 210093,China;School of Mechanics,Civil Engineering and Architecture,Northwestern Polytechnical University,Xi’an 710000,China)

机构地区：[1]南京大学建筑与城市规划学院,江苏南京210093 [2]西北工业大学力学与土木建筑学院,陕西西安710000

出　　处：《山西建筑》2021年第1期180-182,共3页Shanxi Architecture

基　　金：江苏省住房与建设厅科技指导类项目(项目编号:2020ZD55)。

摘　　要：影响高校建筑学专业本科毕业生综合设计潜力的因素较多,涉及面广且关系复杂,具有隐性与显性难以精确划分,定性与定量并存等特点。由于是新兴的研究课题,缺乏长期和大量的数据积累。选择被广泛应用在小样本、非线性及高维数据的SVM神经网络分析系统,利用MATLAB软件和Libsvm工具箱,对建筑学专业本科毕业生综合设计潜力进行建构和综合评价。研究显示该方法较为简洁易用,评价结果也较为客观准确。There are many factors that affect the comprehensive design potential of architectural graduates.These multidimensional and complex factors are difficult to analysis because of their qualitative and quantitative,and implicit and explicit characters.Since lacking of long term and massive data base,Support Vector Machine(SVM)neural network analysis system is used to evaluate the comprehensive design potential of architectural graduates.The SVM is a kind of artificial neural network which is widely used for small samples analysis,nonlinear and high-dimensional data.The SVM evaluation model of the comprehensive design potential is built in MATLAB and Libsvm toolbox.The study shows that the SVM evaluation model is simple and easy to use.The evaluation results are relatively objective and accurate.

关 键 词：综合设计潜力 建筑教育 SVM神经网络 评估 

分 类 号：G642.0[文化科学—高等教育学]