Learning Algorithms for Human-Machine Interfaces

被引:36
作者
Danziger, Zachary [1 ,2 ]
Fishbach, Alon [2 ]
Mussa-Ivaldi, Ferdinando A. [1 ,2 ]
机构
[1] Northwestern Univ, Evanston, IL 60208 USA
[2] Rehabil Inst Chicago, Sensory Motor Performance Program, Chicago, IL 60611 USA
关键词
Adaptive learning; hand posture; human-machine interface; machine learning; BRAIN-COMPUTER INTERFACE; COMMUNICATION; NEURONS;
D O I
10.1109/TBME.2009.2013822
中图分类号
R318 [生物医学工程];
学科分类号
0831 ;
摘要
The goal of this study is to create and examine machine learning algorithms that adapt in a controlled and cadenced way to foster a harmonious learning environment between the user and the controlled device. To evaluate these algorithms, we have developed a simple experimental framework. Subjects wear an instrumented data glove that records finger motions. The high-dimensional glove signals remotely control the joint angles of a simulated planar two-link arm on a computer screen, which is used to acquire targets. A machine learning algorithm was applied to adaptively change the transformation between finger motion and the simulated robot arm. This algorithm was either LMS gradient descent or the Moore-Penrose (MP) pseudoinverse transformation. Both algorithms modified the glove-to-joint angle map so as to reduce the endpoint errors measured in past performance. The MP group performed worse than the control group (subjects not exposed to any machine learning), while the LMS group outperformed the control subjects. However, the LMS subjects failed to achieve better generalization than the control subjects, and after extensive training converged to the same level of performance as the control subjects. These results highlight the limitations of coadaptive learning using only endpoint error reduction.
引用
收藏
页码:1502 / 1511
页数:10
相关论文
共 31 条
[11]   Parallel man-machine training in development of EEG-based cursor control [J].
Kostov, A ;
Polak, M .
IEEE TRANSACTIONS ON REHABILITATION ENGINEERING, 2000, 8 (02) :203-205
[12]   A VARIABLE STEP SIZE LMS ALGORITHM [J].
KWONG, RH ;
JOHNSTON, EW .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 1992, 40 (07) :1633-1642
[13]   A review of classification algorithms for EEG-based brain-computer interfaces [J].
Lotte, F. ;
Congedo, M. ;
Lecuyer, A. ;
Lamarche, F. ;
Arnaldi, B. .
JOURNAL OF NEURAL ENGINEERING, 2007, 4 (02) :R1-R13
[14]   Remapping hand movements in a novel geometrical environment [J].
Mosier, KM ;
Scheidt, RA ;
Acosta, S ;
Mussa-Ivaldi, FA .
JOURNAL OF NEUROPHYSIOLOGY, 2005, 94 (06) :4362-4372
[15]   Spatiotemporal tuning of motor cortical neurons for hand position and velocity [J].
Paninski, L ;
Fellows, MR ;
Hatsopoulos, NG ;
Donoghue, JP .
JOURNAL OF NEUROPHYSIOLOGY, 2004, 91 (01) :515-532
[16]  
Penrose R., 1956, P CAMBRIDGE PHILOS S, V52, P17, DOI [DOI 10.1017/S0305004100030929, 10.1017/s0305004100030929]
[17]   Current trends in Graz Brain-Computer Interface (BCI) research [J].
Pfurtscheller, G ;
Neuper, C ;
Guger, C ;
Harkam, W ;
Ramoser, H ;
Schlögl, A ;
Obermaier, B ;
Pregenzer, M .
IEEE TRANSACTIONS ON REHABILITATION ENGINEERING, 2000, 8 (02) :216-219
[18]  
Santello M, 1998, J NEUROSCI, V18, P10105
[19]   BCI2000: A general-purpose, brain-computer interface (BCI) system [J].
Schalk, G ;
McFarland, DJ ;
Hinterberger, T ;
Birbaumer, N ;
Wolpaw, JR .
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 2004, 51 (06) :1034-1043
[20]   Statistical encoding model for a primary motor cortical brain-machine interface [J].
Shoham, S ;
Paninski, LM ;
Fellows, MR ;
Hatsopoulos, NG ;
Donoghue, JP ;
Normann, RA .
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 2005, 52 (07) :1312-1322