Boostani Reza, Moradi Mohammad Hassan
Amir Kabir University of Technology, Faculty of Biomedical Engineering, Tehran, Iran.
J Neural Eng. 2004 Dec;1(4):212-7. doi: 10.1088/1741-2560/1/4/004. Epub 2004 Nov 17.
High rate classification of imagery tasks is still one of the hot topics among the brain computer interface (BCI) groups. In order to improve this rate, a new approach based on fractal dimension as feature and Adaboost as classifier is presented for five subjects in this paper. To have a comparison, features such as band power, Hjorth parameters along with LDA classifier have been taken into account. Fractal dimension as a feature with Adaboost and LDA can be considered as alternative combinations for BCI applications.
图像任务的高速分类仍然是脑机接口（BCI）研究团队中的热门话题之一。为了提高这一分类速度，本文针对五名受试者提出了一种基于分形维数作为特征、Adaboost作为分类器的新方法。为了进行比较，还考虑了诸如频段功率、Hjorth参数等特征以及线性判别分析（LDA）分类器。分形维数作为一种特征与Adaboost和LDA相结合，可以被视为脑机接口应用的替代组合。
