PL
 |
EN
Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl
Preferencje help
Polish version English version Język
Widoczny [Schowaj] Abstrakt
Liczba wyników

Znaleziono wyników: 7

Liczba wyników na stronie
first rewind previous Strona / 1 next fast forward last
Wyniki wyszukiwania
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 1 next fast forward last
1
Content available Theory II: Deep learning and optimization
100%
Poggio T. , Liao Q.
Bulletin of the Polish Academy of Sciences. Technical Sciences
|
2018
|
tom Vol. 66, nr 6
775--787
EN
The landscape of the empirical risk of overparametrized deep convolutional neural networks (DCNNs) is characterized with a mix of theory and experiments. In part A we show the existence of a large number of global minimizers with zero empirical error (modulo inconsistent equations). The argument which relies on the use of Bezout theorem is rigorous when the RELUs are replaced by a polynomial nonlinearity. We show with simulations that the corresponding polynomial network is indistinguishable from the RELU network. According to Bezout theorem, the global minimizers are degenerate unlike the local minima which in general should be non-degenerate. Further we experimentally analyzed and visualized the landscape of empirical risk of DCNNs on CIFAR-10 dataset. Based on above theoretical and experimental observations, we propose a simple model of the landscape of empirical risk. In part B, we characterize the optimization properties of stochastic gradient descent applied to deep networks. The main claim here consists of theoretical and experimental evidence for the following property of SGD: SGD concentrates in probability – like the classical Langevin equation – on large volume, ”flat” minima, selecting with high probability degenerate minimizers which are typically global minimizers.
2
Content available Theory II: Deep learning and optimization
100%
Poggio T. , Liao Q.
|
|
nr 6
3
Content available Theory I: Deep networks and the curse of dimensionality
100%
Poggio T. , Liao Q.
|
|
nr 6
4
Content available Image encryption combining discrete fractional angular transform with Arnold transform in image bit planes
80%
Zhou Z. , Yu J. , Liao Q. , Gong L.
|
|
tom Vol. 48, nr 2
225--236
EN
A new image encryption algorithm by using a discrete fractional angular transform and Arnold transform in image bit planes is investigated. In the image encryption algorithm, the original image is encrypted by the Arnold transform in image bit planes firstly, and then the resulting image is encrypted by the discrete fractional angular transform further. The key of the image encryption algorithm includes the parameters of the Arnold transform and the order of the discrete fractional angular transform. It is shown that the proposed image encryption algorithm is of high security and strong enough to counteract some conventional image attack manners.
5
Fractions and bioavailability of cadmium and nickel to carrot crops in oasis soi
70%
Wang X. , Nan Z. , Liao Q. , Ding W. , Wu W.
Polish Journal of Environmental Studies
|
2012
|
tom 21
|
nr 6
EN
The Hexi Corridor is the most important area for desert oasis farming in northwestern China. Due to persistent drought and water shortage, sewage irrigation is widely used in this area. Heavy metal pollutants contained in the sewage could remain in the surface layer of agricultural soil and accumulate in plants. Our research used pot experiments to evaluate carrot crop (Daucus carota L.) production, heavy metal uptake, and bioavailability under single cadmium (Cd) or nickel (Ni) contamination and compound (Cd-Ni) contaminations in irrigated desert oasis soil. The results show that Cd existed in the Fe-Mn oxide bound fraction and Ni presented in the residual fraction mainly in original (control) soils. Low concentrations of Cd could promote the growth of carrots, while high concentrations of Cd significantly restrain the growth of the crops. However, Ni had a poisonous effect on the carrots even at the lowest concentrations. There was an antagonistic effect between Cd and Ni in the compound contaminated oasis soils. The bio-concentration factors (BCF) of Cd in carrots were higher than those of Ni, and the BCF of Cd and Ni in single-contaminated soils were higher than those in compound-contaminated soils. Cd and Ni contents in different parts of the carrots were correlated with the exchangeable fraction in contaminated oasis soils, which would cause potential risk to human health through the food chain.
6
Content available remote A Novel Structure Design and Training Algorithm for Quantum Neural Network
61%
Sun J. , Zhang X. , Cao T. , Yang J. , Sun X. , Liao Q.
|
|
tom R. 88, nr 2
315-320
EN
In the structure of original Quantum Neural Network (QNN), only multi-sigmoid transfer function is adopted. Besides that, due to the conflict of the two objective functions in original training algorithm, the training process converges slowly and presents constant variation. In this paper, the QNN with multi-tan-sigmoid transfer function and a novel training algorithm which combines the two objective functions are proposed. Experimental results demonstrate the effectiveness of the structure improvement and the new training algorithm.
PL
W oryginalnym algorytmie kwantowej sieci neuronowej QNN tylko multisigmoidalna funkcja przejścia jest wykorzystywana. W pracy zaprezentowano sieć z multi-tan-sigmoidalną funkcją przejścia z nowym algorytmem uczenia.
7
Effects of soil moisture content on the growth and physiological status of ginger (Zingiber officinale Roscoe)
61%
Li H. , Huang M. , Tan D. , Liao Q. , Zou Y. , Jiang Y.
Acta Physiologiae Plantarum
|
2018
|
tom 40
|
nr 07
first rewind previous Strona / 1 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.