This paper investigates the use of meta-learning to estimate the predictive accuracy of a classifier. We present a scenario where metalearning is seen as a regression task and consider its potential in connection with three strategies of dataset characterization. We show that it is possible to estimate classifier performance with a high degree of confidence and gain knowledge about the classifier through the regression models generated. We exploit the results of the models to predict the ranking of the inducers. We also show that the best strategy for performance estimation is not necessarily the best one for ranking generation.
Abstract. The work shows how a meta-learning technique can be successfully applied to decide when to prune, how much pruning is appropriate and what the best pruning technique is for a given learning task.
Arguably, model selection is one of the major obstacles, and a key once solved, to the widespread use of machine learning/data mining technology in business. Landmarking is a novel and promising metalearning approach to model selection. It uses accuracy estimates from simple and efficient learners to describe tasks and subsequently construct meta-classifiers that predict which one of a set of more elaborate learning algorithms is appropriate for a given problem. Experiments show that landmarking compares favourably with the traditional statistical approach to meta-learning.
0. Neste texto faço um conjunto de observações exploratórias acerca de como lidamos com nossos desejos. Meu pensamento norteador é que nossos desejos são parte de uma arena que é, em sentido amplo, uma arena política e portanto pertence à esfera de nossa soberania em relação, por exemplo, aos ditames da natureza. É porque nossos desejos são politizáveis que nossa libido -o que conduz nossos desejos -pode ser colonizada. Penso que o patriarcado que normatiza a heterossexualidade -o heteropatriarcado -é um regime político acerca dos nossos desejos. Por ser esse regime político alguma coisa que parece estar profundamente enraizado em nossa maneira de pensar (e de escolher como e sobre o que queremos pensar) não achei que poderia abordar o tema senão na forma de explorações em muitas direções. Rondando muitas vezes os condicionamentos de nossas maneiras de pensar, apostei que podemos conseguir pensar ao largo do patriarcado.
This work attempts to articulate a feminist and holist account of our desires and their relation to nature. I consider desires in relation to our bodies, the environment around them and how they respond to the forces within our subjectivities, to society around us and to the evolutionary history of our species. I start out considering the separation of the personal and the political and then move on to make some remarks about nature and the three intertwined ecologies described by Guattari. I suggest that we understand the body as a meeting point for those ecologies and, at the same time, as constituting itself a political platform. I then offer a model of how nature constrains and shape our desire so that no specific message is drawn by our desires from nature. It follows that our desires can be changed as they are not in any sense more natural than the rest of our subjectivity. This change, I claim, is to be thought of in terms of an ecological politics.
This chapter introduces some concepts that help exploring the ontological import of universal logic. It studies the notions of an antilogic and counterlogic associated with each logic and shows some of their properties. It presents the notion of galaxy, as the class of possible worlds compatible with a given logic. We explore some consequences of these developments.
O reconhecimento de uma autoridade (de fatos, de observações compartilhadas, de princípios comumente aceitos etc.) por parte de quem se convence é o que buscamos quando nos engajamos em uma prática de oferecer razões para um pensamento ou uma ação. Parece que o ato do reconhecimento é um elemento constitutivo do estado em que temos (ou aceitamos) razões. Por outro lado, muitas vezes estamos inclinados a dizer que temos ou não temos razão; como se, caso a tenhamos, o reconhecimento da autoridade da nossa razão seja compulsório. Por vezes, quando nos movemos no espaço das razões, temos a impressão de que elas já estão lá, de que elas são independentes do nosso melhor julgamento, e o melhor que temos a fazer é aprender a ficar sensíveis a elas. Neste texto examino o realismo quanto a razões como uma formulação do externalismo e considero as conexões entre externalismo em semântica e em epistemologia. Argumento que um realismo quanto a razões pode fazer uso de uma estratégia davidsoniana para conciliar a idéia de que razões estão no mundo com a concepção destas últimas segundo a qual o reconhecimento lhes é constitutivo.
When we offer reasons for a belief or an action we prompt the audience to recognize authorities (of facts, of shared observations, of commonly held principles etc.). This recognition seems to be constitutive of the very state we find ourselves in when we have or accept reasons. On the other hand, often we are inclined to say that either we have or we don't have reasons for something; we sound as if in case we have reasons, their acceptance is compulsory. Sometimes, when we move in the space of reasons we have the impression that reasons are already there, that they are independent from our best judgment and we should rather learn to be sensitive to their claim. In this paper I consider realism about reasons as a way of stating externalism and I then examine connections between externalism in semantics and in epistemology. I argue that a realism about reasons could make use of a Davidsonian strategy to put together the idea that reasons are in the world with the conception according to which recognition is part of what constitutes a reason
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.