Results 1 - 10 of 486

Expectations, Learning and Macroeconomic Persistence

by Fabio Milani - Journal of Monetary Economics , 2007
"... Abstract. This paper presents an estimated model with learning and provides evidence that learning can improve the …t of popular monetary DSGE models and endogenously generate realistic levels of persistence. The paper starts with an agnostic view, developing a model that nests learning and some of ..."
Abstract - Cited by 103 (7 self) - Add to MetaCart
of the learning gain coe ¢ cient jointly with the ‘deep’parameters of the economy. The empirical results show that when learning replaces rational expectations, the estimated degrees of habits and indexation drop near zero. This …nding suggests that persistence arises in the model economy mainly from expectations

Gaussian processes for machine learning

by Matthias Seeger - International Journal of Neural Systems , 2004
"... Gaussian processes (GPs) are natural generalisations of multivariate Gaussian random variables to infinite (countably or continuous) index sets. GPs have been applied in a large number of fields to a diverse range of ends, and very many deep theoretical analyses of various properties are available. ..."
Abstract - Cited by 92 (14 self) - Add to MetaCart
. This paper gives an introduction to Gaussian processes on a fairly elementary level with special emphasis on characteristics relevant in machine learning. It draws explicit connections to branches such as spline smoothing models and support vector machines in which similar ideas have been investigated

Materials for an exploratory theory of the network society.

by Manuel Castells , Anthony Giddens , Alain Touraine , Anthony Smith , Benjamin Barber , Peter Hall , Roger-Pol Droit , Sophie Watson , Frank Webster , Krishan Kumar , David Lyon , Craig Calhoun , Jeffrey Henderson , Ramon Ramos , Jose E Rodrigues-Ibanez , Jose F Tezanos , Mary Kaldor , Stephen Jones , Christopher Freeman - The British Journal of Sociology , 2000
"... ABSTRACT This article aims at proposing some elements for a grounded theor y of the network society. The network society is the social structure characteristic of the Information Age, as tentatively identi ed by empirical, cross-cultural investigation. It permeates most societies in the world, in v ..."
Abstract - Cited by 122 (0 self) - Add to MetaCart
, international trade of goods and services, advanced business services, multinational production rms and their ancillary networks, communication media, and highly skilled speciality labour. Most jobs are in fact not global, but all economies are under the in uence of the movements of their globalized core

Deep Language and Persistent Culture: Learning to Speak the “Tongue of the Orichas ” in Cuban Santería

by Kristina Wirtz
"... My concern in this paper is to explore how religious practitioners ’ approaches to learning, using, and interpreting Santería’s esoteric ritual language, Lucumí, impact Lucumí’s ritual efficacy as the sacred speech of the deities and ancestors. In particular, I wish to argue that distinct interpreti ..."
Abstract - Add to MetaCart
My concern in this paper is to explore how religious practitioners ’ approaches to learning, using, and interpreting Santería’s esoteric ritual language, Lucumí, impact Lucumí’s ritual efficacy as the sacred speech of the deities and ancestors. In particular, I wish to argue that distinct

An Efficient Learning Procedure for Deep Boltzmann Machines

by Ruslan Salakhutdinov, Geoffrey Hinton , 2010
"... We present a new learning algorithm for Boltzmann Machines that contain many layers of hidden variables. Data-dependent statistics are estimated using a variational approximation that tends to focus on a single mode, and data-independent statistics are estimated using persistent Markov chains. The u ..."
Abstract - Cited by 18 (1 self) - Add to MetaCart
We present a new learning algorithm for Boltzmann Machines that contain many layers of hidden variables. Data-dependent statistics are estimated using a variational approximation that tends to focus on a single mode, and data-independent statistics are estimated using persistent Markov chains

CAPTCHA Recognition with Active Deep Learning

by Fabian Stark, Rudolph Triebel, Daniel Cremers
"... Abstract. CAPTCHAs are automated tests to tell computers and humans apart. They are designed to be easily solvable by humans, but unsolvable by machines. With Convolutional Neural Networks these tests can also be solved automatically. However, the strength of CNNs relies on the training data that th ..."
Abstract - Add to MetaCart
that the classifier is learnt on and especially on the size of the training set. Hence, it is intractable to solve the problem with CNNs in case of insufficient training data. We propose an Active Deep Learning strategy that makes use of the ability to gain new training data for free without any human intervention

Boosted Backpropagation Learning for Training Deep Modular Networks

by Alexander Grubb, J. Andrew Bagnell
"... Divide-and-conquer is key to building sophisticated learning machines: hard problems are solved by composing a network of modules that solve simpler problems (LeCun et al., 1998; Rohde, 2002; Bradley, 2009). Many such existing systems rely on learning algorithms which are based on simple parametric ..."
Abstract - Cited by 4 (1 self) - Add to MetaCart
Divide-and-conquer is key to building sophisticated learning machines: hard problems are solved by composing a network of modules that solve simpler problems (LeCun et al., 1998; Rohde, 2002; Bradley, 2009). Many such existing systems rely on learning algorithms which are based on simple parametric

Deep Submodular Functions: Definitions & Learning

by Brian Dolhansky , Jeff Bilmes , <bilmes@uw Edu> , Engineering
"... Abstract We propose and study a new class of submodular functions called deep submodular functions (DSFs). We define DSFs and situate them within the broader context of classes of submodular functions in relationship both to various matroid ranks and sums of concave composed with modular functions ..."
Abstract - Add to MetaCart
(SCMs). Notably, we find that DSFs constitute a strictly broader class than SCMs, thus motivating their use, but that they do not comprise all submodular functions. Interestingly, some DSFs can be seen as special cases of certain deep neural networks (DNNs), hence the name. Finally, we provide a method

A Deep Embedding Model for Co-occurrence Learning

by Yelong Shen , Ruoming Jin , Jianshu Chen , Xiaodong He , Jianfeng Gao , Li Deng
"... Abstract-Co-occurrence Data is a common and important information source in many areas, such as the word co-occurrence in the sentences, friends co-occurrence in social networks and products co-occurrence in commercial transaction data, etc, which contains rich correlation and clustering informatio ..."
Abstract - Add to MetaCart
, we apply the maximum pseudolikelihood method to learn DEM. In consequence, the developed model and its learning method naturally avoid the above difficulties and can be easily used to compute the conditional probability in prediction. Interestingly, our method is equivalent to learning a special

Learning Factored Representations in a Deep Mixture of Experts

by David Eigen, Marc’aurelio Ranzato Ilya Sutskever, Google Inc
"... Mixtures of Experts combine the outputs of several “expert ” networks, each of which specializes in a different part of the input space. This is achieved by train-ing a “gating ” network that maps each input to a distribution over the experts. Such models show promise for building larger networks th ..."
Abstract - Add to MetaCart
input with a combination of experts at each layer, yet maintains a modest model size. On a randomly translated version of the MNIST dataset, we find that the Deep Mixture of Experts automatically learns to develop location-dependent (“where”) experts at the first layer, and class-specific (“what
Results 1 - 10 of 486