1,087 research outputs found
Quantity makes quality: learning with partial views
No full textIn many real world applications, the number of examples to learn from is plentiful, but we can only obtain limited information on each individual example. We study the possibilities of efficient, provably correct, large-scale learning in such settings. The main theme we would like to establish is that large amounts of examples can compensate for the lack of full information on each individual example. The type of partial information we consider can be due to inherent noise or from constraints on the type of interaction with the data source. In particular, we describe and analyze algorithms for budgeted learning, in which the learner can only view a few attributes of each training example (Cesa-Bianchi, Shalev-Shwartz, and Shamir 2010a; 2010c), and algorithms for learning kernel-based predictors, when individual examples are corrupted by random noise (Cesa-Bianchi, Shalev-Shwartz, and Shamir 2010b)
Online learning with switching costs and other adaptive adversaries
No full textWe study the power of different types of adaptive (nonoblivious) adversaries in the setting of prediction with expert advice, under both full-information and bandit feedback. We measure the player's performance using a new notion of regret, also known as policy regret, which better captures the adversary's adaptiveness to the player's behavior. In a setting where losses are allowed to drift, we characterize-in a nearly complete manner-the power of adaptive adversaries with bounded memories and switching costs. In particular, we show that with switching costs, the attainable rate with bandit feedback is ⊖∼(T2/3). Interestingly, this rate is significantly worse than the ⊖(√T) rate attainable with switching costs in the full-information case. Via a novel reduction from experts to bandits, we also show that a bounded memory adversary can force ⊖∼(T2/3) regret even in the full information case, proving that switching costs are easier to control than bounded memory adversaries. Our lower bounds rely on a new stochastic adversary strategy that generates loss processes with strong dependencies
Early literacy interventions using ICT in children with SLI
No full textContains fulltext :
121119.pdf (Publisher’s version ) (Open Access
Bandit Regret Scaling with the Effective Loss Range
No full textWe study how the regret guarantees of nonstochastic multi-armed bandits can be improved, if the effective range of the losses in each round is small (for example, the maximal difference between two losses or in a given round). Despite a recent impossibility result, we show how this can be made possible under certain mild additional assumptions, such as availability of rough estimates of the losses, or knowledge of the loss of a single, possibly unspecified arm, at the end of each round. Along the way, we develop a novel technique which might be of independent interest, to convert any multi-armed bandit algorithm with regret depending on the loss range, to an algorithm with regret depending only on the effective range, while attaining better regret bounds than existing approaches
Efficient transductive online learning via randomized rounding
No full textMost traditional online learning algorithms are based on variants of mirror descent or follow-the-leader. In this paper, we present an online algorithm based on a completely different approach, tailored for transductive settings, which combines "random playout" and randomized rounding of loss subgradients.
As an application of our approach, we present the first computationally
efficient online algorithm for collaborative filtering with trace-norm
constrained matrices. As a second application, we solve an open question
linking batch learning and transductive online learning
Efficient learning with partially observed attributes
No full textWe describe and analyze efficient algorithms for learning a linear predictor from examples when the learner can only view a few attributes of each training example. This is the case, for instance, in medical research, where each patient participating in the experiment is only willing to go through a small number of tests. Our analysis bounds the number of additional examples sufficient to compen-sate for the lack of full information on each training example. We demonstrate the efficiency of our algorithms by showing that when running on digit recognition data, they obtain a high prediction accuracy even when the learner gets to see only four pixels of each image. Copyright 2010 by the author(s)/owner(s)
estudo de caso: Shamir Optical
No full textEsta versão não contém as críticas e sugestões dos elementos do júriO presente trabalho enquadra-se no âmbito da realização do relatório de estágio para a obtenção do grau de mestre em Assessoria de Administração e tem como título a Comunicação Externa em Business-to-Business (Estudo de Caso: Shamir Optical).
Este relatório visa o estudo da importância da comunicação externa de organizações que praticam business-to-business, e analisar as suas ferramentas e componentes. A escolha do tema deve-se ao facto de se verificar uma preocupação crescente em relação à comunicação externa a nível global, abrangendo também as organizações de business-to-business. Nestas organizações, os processos comunicacionais utilizados não foram, até ao momento, alvo de muitos estudos ao contrário do que acontece nas organizações de business-to-consumer.
Para a sua concretização, além de um estágio profissional, foi realizado um estudo de caso. Este estudo teve o intuito de analisar a comunicação externa na empresa de business-to-business do setor ótico Shamir Optical.
Foram realizadas pesquisas e estudos no ambito do tema abordado com o objetivo de descobrir e analisar aspetos peculiares da comunicação externa das organizações de business-to-business. Para isso, foram convergidos dados recolhidos através de uma entrevista semiestruturada, observação participante e análise documental.
Das informações recolhidas extrapolamos que a comunicação externa tem vindo a assumir uma importância crescente no panorama organizacional global, inclusive nas organizações de business-to-business, como é o caso da Shamir Optical sendo cada vez mais recursos redirecionados para esta área.This paper is enclosed in the ambit of the internship report to obtain a master's degree in Administrative Assistence and is titled as External Communication in Business-to-Business (Case Study: Shamir Optical).
This report aims to study the importance of the external communication, of organizations involved in business-to-business, and analyze the tools and components used. The choice of the research topic is owing to fact that there is a growing concern about global external communication, also incluing business-to-business organizations. In these organizations, the communication processes used were not so far, the subject of lots of studies, opposing what happens with business-to-consumer subject.
An internship and a case study was done to accomplish the goals. This study aimed to analyze the external communication in business-to-business organization Shamir Optical.
Researches and studies were made on the subject in order to discover and analyze the atypical aspects of the external communication of business-to-business organizations. In order to accomplish this, was collected data through an interview, participant observation and documental investigation.
We extrapolated from the information collected that external communication has become increasingly important in the global organizational landscape, including the business-to-business organizations, such the case of Shamir Optical increasingly resources sent to this area
Classification of large acoustic datasets using machine learning and crowdsourcing : application to whale calls
No full textA.M.v.B.B. acknowledges support from the Sea Mammal Research Unit at the University of St. Andrews (Professor Ian Boyd) and Woods Hole Oceanographic Institution for the Whale FM project. P.T. received funding from the Marine Alliance for Science and Technology for Scotland (MASTS).Vocal communication is a primary communication method of killer and pilot whales, and is used for transmitting a broad range of messages and information for short and long distance. The large variation in call types of these species makes it challenging to categorize them. In this study, sounds recorded by audio sensors carried by ten killer whales and eight pilot whales close to the coasts of Norway, Iceland, and the Bahamas were analyzed using computer methods and citizen scientists as part of the Whale FM project. Results show that the computer analysis automatically separated the killer whales into Icelandic and Norwegian whales, and the pilot whales were separated into Norwegian long-finned and Bahamas short-finned pilot whales, showing that at least some whales from these two locations have different acoustic repertoires that can be sensed by the computer analysis. The citizen science analysis was also able to separate the whales to locations by their sounds, but the separation was somewhat less accurate compared to the computer method.Peer reviewe
- …
