This paper presents the prior PAC-Bayes bound and explores its capabilities as a tool to provide tight predictions of SVMs' generalization. The computation of the bound involves estimating a prior of the distribut...
详细信息
This paper presents the prior PAC-Bayes bound and explores its capabilities as a tool to provide tight predictions of SVMs' generalization. The computation of the bound involves estimating a prior of the distribution of classifiers from the available data, and then manipulating this prior in the usual PAC-Bayes generalization bound. We explore two alternatives: to learn the prior from a separate data set, or to consider an expectation prior that does not need this separate data set. The prior PAC-Bayes bound motivates two SVM-like classification algorithms, prior SVM and ν-prior SVM, whose regularization term pushes towards the minimization of the prior PAC-Bayes bound. The experimental work illustrates that the new bounds can be significantly tighter than the original PAC-Bayes bound when applied to SVMs, and among them the combination of the prior PAC-Bayes bound and the prior SVM algorithm gives the tightest bound.
In this paper, we propose a higher order safe ambients calculus. In this setting, we propose an extended labelled transition system and an extended labelled bisimulation for this calculus. We also give the reduction b...
详细信息
Criterion, standard and policy of Korean language have played a very important role as Autonomy regulations in many aspects, especially input standards of computer using Korean language. So it is very necessary to inv...
详细信息
The objective of semantic segmentation in microscopic images is to extract the cellular, nuclear or tissue components. This problem is challenging due to the large variations of these components features (size, shape,...
详细信息
A network code assisted HARQ system considering uplink (UL) transmission with multiple mobile stations (MSs) cooperating with a relay-station (RS) is proposed. The RS is assumed to successfully monitor and decode the ...
详细信息
Feature extraction is an important step in pattern classification and speech recognition. Extracted features should discriminate classes from each other while being robust to the environmental conditions such as noise...
详细信息
Text classification is the key technology for topic tracking, and vector space model (VSM) is one of the most simple and effective model for topics representation. Feature selection algorithm in VSM is an important me...
详细信息
For services that have similar functionalities, if they are published by different cloud platforms, it is a challenge to evaluate them, for satisfying different end users' personal preferences. In view of this cha...
详细信息
Users usually have different prospective even they input a same keyword to search Web services. It is a challenge to personalize web service search engine as more and more keyword-like Web services becoming available ...
详细信息
The potential use of molecular computation in attacking the Data Encryption Standard (DES) is already known, but the used computing models are not autonomous and require many tedious laboratory steps to execute. In th...
详细信息
暂无评论