CMU-Repository: ftp://ftp.cs.cmu.edu/afs/cs.cmu.edu/ project/connect/bench/. WWW Repository, Carnegie Mellon University (1993) Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Evolution of Cubic Spline Activation Functions for.. - Mayer, Schwaiger (2001) (Correct)
....and the test set is made of 10,000 equally spaced points covering the complete unit square. The task in the Two Spirals problem is to discriminate between two sets of points which lie on two distinct spirals in the plane. These spirals coil three times around the origin and around one another [10]. Training, validation, and test set comprise 194 patterns each. The basic ANN structure is de ned by two input neurons and two output neurons (one for each spiral, winner takes all) For these classi cation problems the model error in the tness function is simply given by Em = ev nv , ....
CMU-Repository: ftp://ftp.cs.cmu.edu/afs/cs.cmu.edu/ project/connect/bench/. WWW Repository, Carnegie Mellon University (1993)
Differentiation of Neuron Types by Evolving Activation.. - Mayer, Schwaiger (2002) (Correct)
....and the test set is made of 10,000 equally spaced points covering the complete unit square. The task in the Two Spirals problem is to discriminate between two sets of points which lie on two distinct spirals in the plane. These spirals coil three times around the origin and around one another [2]. Training, validation, and test set comprise 194 patterns each. The basic ANN structure is de ned by two input neurons and two output neurons (one for each spiral, winner takes all) For these classi cation problems the model error in the tness function is simply given by Em = ev nv , where ....
....problems, we also present results on a time series prediction task. The Mackey Glass time series is generated by a di erential delay equation. The four input neurons represent the points in time (x(t 18) x(t 12) x(t 6) x(t) and the network is trained to predict the future value x(t 84) [2]. The original benchmark data is partitioned in a training set (3000 patterns, t = 200 3200) and a test set (500 patterns t = 5000 5500) We used the rst 1000 patterns of the training set and the next 2000 patterns for the validation set (test set unchanged) For all experiments the identity AF ....
CMU-Repository. ftp://ftp.cs.cmu.edu/afs/cs.cmu.edu/ project/connect/bench/. WWW Repository, Carnegie Mellon University, 1993.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC