Improving the Performance of Radial BasisFunction

نویسندگان

  • Dietrich Wettschereck
  • Thomas Dietterich
  • J E Moody
  • S J Hanson
  • R P Lippmann
چکیده

Three methods for improving the performance of (gaussian) radial basis function (RBF) networks were tested on the NETtalk task. In RBF, a new example is classiied by computing its Euclidean distance to a set of centers chosen by unsupervised methods. The application of supervised learning to learn a non-Euclidean distance metric was found to reduce the error rate of RBF networks, while supervised learning of each center's variance resulted in inferior performance. The best improvement in accuracy was achieved by networks called generalized radial basis function (GRBF) networks. In GRBF, the center locations are determined by supervised learning. After training on 1000 words, RBF classiies 56.5% of letters correct, while GRBF scores 73.4% letters correct (on a separate test set). From these and other experiments, we conclude that supervised learning of center locations can be very important for radial basis function learning.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

A chaos tracker applied to non-coherent detection in chaos-based digital communication systems

In chaos-based communication systems, parameter variation in the chaos generator and additive channel noise represent two practical problems that are hard to sol*. For chaos-based digital communication, non-coherent detection has the advantage that the receiver does not need to reproduce the same chaos basis function that has been generated in the transmitter. SUC h reproduction typically requi...

متن کامل

Development of a Non-Radial Network Model to Evaluate the Performance of a Multi-Stage Sustainable Supply Chain

 Abstract:The purpose of this paper is to present a new model of non-radial data envelopment analysis that is able to evaluate the systems one of these complete networks is supply chain of cement industry. In this paper, using a non-radial model in data envelopment analysis, a model with a network structure that can assess the sustainable supply chain of strategic industries is evaluated....

متن کامل

The effect of inclined radial flow in proton exchange membrane fuel cells performance

Computational fluid dynamics analysis was employed to investigate the radial flow field patterns of proton exchange membrane fuel cells (PEMFC) with different channel geometries at high operating current densities. 3D, non-isothermal was used with single straight channel geometry. Our study showed that new generation of fuel cells with circle stack with the same active area and inlet area gave ...

متن کامل

Improving Accuracy of DGPS Correction Prediction in Position Domain using Radial Basis Function Neural Network Trained by PSO Algorithm

Differential Global Positioning System (DGPS) provides differential corrections for a GPS receiver in order to improve the navigation solution accuracy. DGPS position signals are accurate, but very slow updates. Improving DGPS corrections prediction accuracy has received considerable attention in past decades. In this research work, the Neural Network (NN) based on the Gaussian Radial Basis Fun...

متن کامل

Optimal reconfiguration of radial distribution system with the aim of reducing losses and improving voltage profiles using the improved lightning search algorithm

In this paper, a modified version of the lightning search algorithm is proposed in order to find the optimal reconfiguration of the switches and locate and determine the optimal capacity of distributed generation sources in the distribution feeder. The main optimization goals are to reduce ohmic losses and voltage deviations in the standard 33-bus and 94-node IEEE feeders. The simulation result...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1992