✦ LIBER ✦

Learning and Approximation Capabilities of Adaptive Spline Activation Function Neural Networks

✍ Scribed by Lorenzo Vecci; Francesco Piazza; Aurelio Uncini

Publisher: Elsevier Science
Year: 1998
Tongue: English
Weight: 297 KB
Volume: 11
Category: Article
ISSN: 0893-6080
DOI: 10.1016/s0893-6080(97)00118-4

No coin nor oath required. For personal study only.

✦ Synopsis

In this paper, we study the theoretical properties of a new kind of artificial neural network, which is able to adapt its activation functions by varying the control points of a Catmull-Rom cubic spline. Most of all, we are interested in generalization capability, and we can show that our architecture presents several advantages. First of all, it can be seen as a sub-optimal realization of the additive spline based model obtained by the reguralization theory. Besides, simulations confirm that the special learning mechanism allows to use in a very effective way the network's free parameters, keeping their total number at lower values than in networks with sigmoidal activation functions. Other notable properties are a shorter training time and a reduced hardware complexity, due to the surplus in the number of neurons.

📜 SIMILAR VOLUMES

A neural network capable of learning and

A neural network capable of learning and inference for visual pattern recognition

✍ Ho J Kim; Hyun S Yang 📂 Article 📅 1994 🏛 Elsevier Science 🌐 English ⚖ 869 KB

Numerical solution of stiff and convecti

Numerical solution of stiff and convection-diffusion equations using adaptive spline function approximation

✍ M.K. Jain; Tariq Aziz 📂 Article 📅 1983 🏛 Elsevier Science 🌐 English ⚖ 535 KB

Neural network approximation of continuo

Neural network approximation of continuous functionals and continuous functions on compactifications

✍ M.B. Stinchcombe 📂 Article 📅 1999 🏛 Elsevier Science 🌐 English ⚖ 164 KB

This article characterizes the set of activation functions, bounded or unbounded, that allow feedforward network approximation of the continuous functions on the classic two-point compactification of R 1 . The characterization fails when the set of targets are continuous functions on the classic com

Properties and performance of orthogonal

Properties and performance of orthogonal neural network in function approximation

✍ Chieh F. Sher; Ching-Shiow Tseng; Chen-San Chen 📂 Article 📅 2001 🏛 John Wiley and Sons 🌐 English ⚖ 216 KB

Backpropagation neural network has been applied successfully to solving uncertain problems in many fields. However, unsolved drawbacks still exist such as the problems of local minimum, slow convergence speed, and the determination of initial weights and the number of processing elements. In this pa

Selection of network and learning parame

Selection of network and learning parameters for an adaptive neural robotic control scheme

✍ N. Sundararajan; Leonard Chin; Yip Kim San 📂 Article 📅 1993 🏛 Elsevier Science 🌐 English ⚖ 990 KB

Approximation of a function and its deri

Approximation of a function and its derivative with a neural network

✍ Pierre Cardaliaguet; Guillaume Euvrard 📂 Article 📅 1992 🏛 Elsevier Science 🌐 English ⚖ 998 KB

This paper deals with the approximation of both a function and its derivative by feedforward neural networks. We propose an explicit formula of approximation which is noise resistant and can be easily modified with the patterns. We apply these results to approach a function defined implicitly, which