Neural Network Trainer with Second Order Learning Algorithms

This paper describes a method of linearizing the nonlinear characteristics of many sensors using an embedded neural network. The proposed method allows for complex neural networks with very powerful architectures to be embedded on a very inexpensive 8-bit microcontroller. In order to accomplish this unique training software was developed as well as a cross compiler. The Neuron by Neuron process was as developed in assembly language to allow the fastest and shortest code on the embedded system. The embedded neural network also required an accurate approximation for hyperbolic tangent to be used as the neuron activation function.This process was then demonstrated on a robotic arm kinematics problem.

show abstract

“…It has been shown that fully connected networks are easier to train and produce better results with smaller networks [13][14][15][16].…”

Section: Abitrarily Connected Networkmentioning

confidence: 99%

Compensation of Sensors Nonlinearity with Neural Networks

Cotton

Wilamowski

2010

2010 24th IEEE International Conference on Advanced Information Networking and Applications

Self Cite

View full text Add to dashboard Cite

show abstract

“…15 LM algorithm was a powerful algorithm in improving the convergence speed of the ANN with multi-layer perceptron (MLP) architectures. 16 It was a good combination of Newton's method and steepest descent.…”

Section: Introductionmentioning

confidence: 99%

Image correction for cone-beam computed tomography simulator using neural network corrector

Chen

Hsu

Chen

et al. 2017

Advances in Mechanical Engineering

View full text Add to dashboard Cite

In this article, a neural network corrector is proposed to correct the image shift, yielding the degradation of threedimensional image reconstruction, for each slice captured by cone-beam computed tomography simulator. There are 3 degrees of freedom in tube module of simulator; the central point of tube module should be aligned with the central point of detector module to guarantee the accurate image projection. However, the mechanism manufacturing and assembling tolerance will let the above aim cannot be met. Here, a standard kit is made to measure the image shift by 1°s tep from 210°to 10°. The measure data will be the input training data of proposed neural network corrector, and the corrected translation position will be the output of neural network corrector. The Levenberg-Marquardt learning algorithm adjusts the connected weights and biases of the neural network using a supervised gradient descent method, such that the defined error function can be minimized. To avoid the problem of overfitting and improve the generalized ability of the neural network, Bayesian regularization is added to the Levenberg-Marquardt learning algorithm. After the training of neural network corrector, the different target position commands are fed into the neural network corrector. Then, the corrected data from neural network corrector are fed to be the new position command to verify the image correction performance. Moreover, a phantom kit is made to check the corrected performance of the neural network corrector. Finally, the experimental results verify that the image shift can be reduced by the neural network corrector.

show abstract

“…Generally, calculation of the Hessian matrix requires a lot of calculation power. However, it is possible to find new methods for Hessian calculation in the second order training algorithms [5], [6]. In the other solution only signs of the cost function first derivatives versus weights are used.…”

Section: Introductionmentioning

confidence: 99%