In this paper the limitations of the sensitivity analysis method for feedforward neural networks in the cases of dependent input variables are discussed. First, it is explained that in such cases there can be many fun...
详细信息
ISBN:
(纸本)1424400716
In this paper the limitations of the sensitivity analysis method for feedforward neural networks in the cases of dependent input variables are discussed. First, it is explained that in such cases there can be many functions implemented by neural networks that will accurately approximate training patterns. Then it is pointed out that many of these functions do not allow for proper estimation of the inputs importance using the sensitivity analysis method for neural networks. These two facts are demonstrated to be the reason why one can not completely rely upon the results of this method, when evaluating a real importance of inputs. Examples with graphs visualizing the discussed phenomena are presented. Finally, general conclusions about overall usefulness of the method are introduced.
作者:
Prof. Jian-Xin XuProf. Leonid FridmanDepartment of Electrical and Computer Eng. National University of Singapore 4 Engineering Drive 3 Singapore 117576 Tel +65 6874-2566
Fax +65 6779-1103 Dr Jian-Xin Xu received his Bachelor degree from Zhejiang University
China in 1982. He attended the University of Tokyo Japan where he received his Master's and Ph.D. degrees in 1986 and 1989 respectively. All his degrees are in Electrical Engineering. He worked for one year in the Hitachi research Laboratory Japan and for more than one year in Ohio State University U.S.A. as a Visiting Scholar. In 1991 he joined the National University of Singapore and is currently an associate professor in the Department of Electrical Engineering. His research interests lie in the fields of learning control variable structure control fuzzy logic control discontinuous signal processing and applications to motion control and process control problems. He is the associate editor of Asian Journal of Control member of TC on variable structure systems and sliding mode control of IEEE Control Systems Society and a senior member of IEEE. He has produced more than 90 peer-refereed journal papers near 160 technical papers in conference proceedings and authored/edited 4 books. Division de Estudios de Posgrado Facultad de Ingenieria National Autonomous University of Mexico DEP-FI
UNAM Edificio “A” Circuito Exterior Ciudad Universitaria A. P. 70–256 C.P.04510 Mexico D.F. Mexico Tel +52 55 56223014 Fax +52 55 56161719 Dr. Leonid M. Fridman received his M.S in mathematics from Kuibyshev (Samara) State University
Russia Ph.D. in Applied Mathematics from Institute of Control Science (Moscow) and Dr. of Science degrees in Control Science from Moscow State University of Mathematics and Electronics in 1976 1988 and 1998 respectively. In 1976–1999 Dr. Fridman was with the Department of Mathematics at the Samara State Architecture and Civil Engineering Academy Samara Russia. In 2000–2002 he was with the Department of Postgraduate Study and Investigations at the Chihuahu
An investigation of the influence of organic vapor (ethanol, methanol and acetone) on the characteristics of simple-structure sensors fabricated from single-walled carbon nanotube (SWCNT) thin films is reported. SWCNT...
详细信息
This paper presents the design and results of first measurements of a fast binary readout architecture ASIC aimed for digital X-ray imaging. The ASIC called DEDIX includes 64 readout channels, a number of DACs for dif...
详细信息
The aim of this paper is to propose a new adaptive technique to improve the performance of 3G Ad-Hoc networks based on CDMA2000 in fading channels. The RAKE receiver combined with channel searchers provide improved es...
详细信息
The aim of this paper is to propose a new adaptive technique to improve the performance of 3G Ad-Hoc networks based on CDMA2000 in fading channels. The RAKE receiver combined with channel searchers provide improved estimation of the Time of Arrival and Amplitude of Arrival of the strongest received signal. Ad-Hoc networks based on the Grid protocol are considered for the analysis and simulation. A novel two-dimension figure of merit is suggested to describe the channel effects on delay and amplitude estimation. Errors in delay are corrected by adaptively increasing the number of RAKE fingers while the amplitude errors are corrected through adaptive power control.
We present the development of a 64 channel readout ASIC called DEDIX for high count rate position-sensitive measurements using semiconductor detectors. The ASIC is designed in 0.35 μm CMOS process and its total area ...
详细信息
The spatial and temporal non-uniformities in infrared focal plane array (IRFPA), which is called as fixed pattern noise, results a slowly varying pattern on the thermal image thus degrading the resolving capabilities ...
详细信息
The spatial and temporal non-uniformities in infrared focal plane array (IRFPA), which is called as fixed pattern noise, results a slowly varying pattern on the thermal image thus degrading the resolving capabilities of thermal imaging system considerably. An adaptive filter based technique is presented to correct fixed pattern noise. This method uses kalman filter and corrects both the additive and multiplicative types of fixed pattern noise thus eliminating the need of halting the operation of thermal camera for calibration. A FPGA based approach is presented for implementing the proposed method. The result obtained and the performance parameter of the algorithm is presented.
This paper describes a novel straight and circular road driving control scheme for electric power assisted wheelchairs. "Electric power assisted wheelchair" which assists the driving force by electric motors...
详细信息
This paper describes a novel straight and circular road driving control scheme for electric power assisted wheelchairs. "Electric power assisted wheelchair" which assists the driving force by electric motors is expected to be widely used as a mobility support system for elderly people and disabled people, however, the performance of the straight and circular road driving must be further improved because the two wheels drive independently. This paper proposes a novel driving control scheme based on fuzzy algorithm to realize the stable and reliable driving on straight and circular roads. The suitable assisted torque of the right and left wheels is determined by fuzzy algorithm based on the posture angular velocity of the wheelchair and the human input torque proportion of the right and left wheels. Some experiments on the practical roads show the effectiveness of the proposed control system
This paper describes a novel step climbing control scheme for electric power assisted wheelchairs. "Electric power assisted wheelchair" which assists the driving force by electric motors is expected to be wi...
详细信息
This paper describes a novel step climbing control scheme for electric power assisted wheelchairs. "Electric power assisted wheelchair" which assists the driving force by electric motors is expected to be widely used as a mobility support system for elderly people and disabled people. This research aims at realizing the safe and reliable step climbing motion control of power assisted wheelchairs because power assisted wheelchairs are in danger of overturning backward when raising the front wheels due to the excessive driving torque and of stopping in front of steps when the rear wheels try to climb over steps due to the insufficient driving torque. This paper proposes a novel step climbing control scheme for power assisted wheelchairs based on driving mode switching. The first driving mode, "front wheel raising mode", applies torque control system with the variable assistance ratio to prevent the dangerous overturning. The second driving mode, "rear wheel climbing mode", applies position control system based on the polynomial trajectory to surely climb over steps and stop at the suitable position. The mode switching control based on these two driving modes can generate the safe and reliable step climbing motions. Some experiments show the effectiveness of the proposed control system
A transmultiplexer combines several images into one image to be sent through a single communication channel is presented. The paper presents the implementation of such system in Field Programmable Gate Arrays. The pro...
详细信息
ISBN:
(纸本)8392117263
A transmultiplexer combines several images into one image to be sent through a single communication channel is presented. The paper presents the implementation of such system in Field Programmable Gate Arrays. The proposed algorithm ensures parallel data flow and calculations for very fast signal processing.
暂无评论