An Adaptive Momentum Back Propagation (AMBP)

Drago, Gian Paolo; Morando, M.; Ridella, Sandro

doi:10.1007/bf01414646

Cited by 17 publications

(10 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…0 implies jgj ! 1 in (7) or (9). Thus a finite gain is associated with a distance threshold, marking a "neutral" stripe running along partition boundaries.…”

Section: Theorem 2 (-Cbp's Embed Vq Network)mentioning

confidence: 98%

Circular backpropagation networks embed vector quantization

Ridella

Rovetta

Zunino

1999

IEEE Trans. Neural Netw.

Self Cite

View full text Add to dashboard Cite

This letter proves the equivalence between vector quantization (VQ) classifiers and circular backpropagation (CBP) networks. The calibrated prototypes for a VQ schema can be plugged in a CBP feedforward structure having the same number of hidden neurons and featuring the same mapping. The letter describes how to exploit such equivalence by using VQ prototypes to perform a meaningful initialization for BP optimization. The approach effectiveness was tested considering a real classification problem (NIST handwritten digits).

show abstract

“…0 implies jgj ! 1 in (7) or (9). Thus a finite gain is associated with a distance threshold, marking a "neutral" stripe running along partition boundaries.…”

Section: Theorem 2 (-Cbp's Embed Vq Network)mentioning

confidence: 98%

Circular backpropagation networks embed vector quantization

Ridella

Rovetta

Zunino

1999

IEEE Trans. Neural Netw.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The value of the learning rate has been made adpative through the algorithm by Vogl et al [12], while, as regards the value of the moment, the expression we proposed by Drago and Ridella [13] has been used:…”

Section: Iap Weight Initialisationmentioning

confidence: 99%

Possibility and Necessity Pattern Classification using an Interval Arithmetic Perceptron

Drago

Ridella

1999

NCA

Self Cite

View full text Add to dashboard Cite

In the work presented in this paper, an Interval Arithmetic Perceptron (IAP) is used to detect the region in the input space to which an uncertainty decision should be appropriately associated. This region may be originated both by sub-regions which are not represented in the training set, and by subregions where the probabilities of the two classes are very similar. To train the IAP, an algorithm will be presented which in particular is able detect the two certainty regions and the uncertainty one. From the interval weights thus obtained, a confidence interval of the probability will also be evaluated. The algorithm has been used for studying a simple artificial problem and two real-world applications, the Iris and Breast Cancer databases. Regarding the latter application in particular, a statistical analysis of the results is presented, together with a discussion of the possible alternative classifications of the patterns attributed to the uncertainty region.

show abstract

“…Bu sakıncaları gidermek üzere, denklem (9) ile verilen öğrenme algoritmasına momentum terimi ekleme, öğrenme oranini ve momentum katsayısuu çeşitli yöntemlerle uyarlama ve eşlenik eğim yöntemi ile ağırlıkları yenileme gibi öğrenme algoritmaları geliştirilmiştir [15][16][17]. Denklem (10) a momentim terimi eklenerek YSA agıriıklanna uygulanacak düzeltme…”

unclassified

“…Sonuçta YSA' mn eğitimi hızlanırken yöresel en azdan kurtulma olasılığı da artacaktır. Geliştirilmiş çeşitli geriye yayılım algoritmaları ile birlikte geri beslemeli YSA nın eğitiminde kullanılan dinamik geriye yayılım algoritmalanndan bazılan kaynak [14][15][16][17][18] …”

unclassified

See 1 more Smart Citation

Self-tuning adaptive neurocontroller for brushless DC motors

Albostan¹,

Gökbulut²

2001

International Journal of Electronics

View full text Add to dashboard Cite

Motors. Nonîinear and unkno~wn motor dynamics are identified by îising multüayer neuraî / network and the control inpuifor the motor is derived fi-om the identifıed model. The effecî of the loadlorqııe on tha control syslem is damped byfıllering the coMrol input. Simulation and experimenîal results show thaî the self-tunmg adaptive neurocontrol has a good tracking performance bul needs an adaplive filler and a paraile! PI controller design in the case of disîurbances.

show abstract

An Adaptive Momentum Back Propagation (AMBP)

Cited by 17 publications

References 12 publications

Circular backpropagation networks embed vector quantization

Circular backpropagation networks embed vector quantization

Possibility and Necessity Pattern Classification using an Interval Arithmetic Perceptron

Self-tuning adaptive neurocontroller for brushless DC motors

Contact Info

Product

Resources

About