Statistical learning theory

About: Statistical learning theory is a research topic. Over the lifetime, 1618 publications have been published within this topic receiving 158033 citations.

Distribution grid reconfiguration based on extreme learning machine

Abstract: To minimize the active power loss of distribution grid reconfiguration,a neural network reconfiguration model based on the extreme learning machine is proposed,which reflects the nonlinear relationship between the load pattern and the switch state of distribution grid.Having simple network structure and fast training speed,the model takes the load pattern as its input and outputs the switch states to reconfigure the distribution grid with minimum active power loss.The structural risk minimization rule of the statistical learning theory is introduced into the extreme learning machine based on the empirical risk minimization to minimize the empirical risk and confidence interval.The actual risk is thus minimized and the expectation error is decreased.Simulative research is carried out for two typical cases of distribution network reconfiguration with different reconfiguration models:support vector machine,BP neural network and extreme learning machine.Results show that the proposed model has both better generalization performance and faster training speed.

Research on multi-classification based on support vector machine

Abstract: Most of the existing methods are based on traditional statistics,which provides that conclusion only for the situation where sample size is tending to infinity.So they may not work well in practical case with limi-ted samples and easily lead to the problem of overfilling.This paper introduced the support vector machine(SVM) based on the theory of traditional statistics.This method can solve small-sample learning problems better by using experiential risk minimization(ERM) in place of structural risk minimization(SRM).Moreover,this theory can change the problem in non-linearity space to that in the linearity space in order to reduce the algorithm complexity by using the kernel function idea.It studies some relational contents including the optimization algorithm and the solution to multi-classification.Finally,through an example,it shows that the pro-posed method is effective and feasible.

The statistical evidence for negative transfer in part-whole free recall

Abstract: A series of studies on part-whole free recall led to the conclusion that learning part of a list before learning the entire list produces negative transfer late in learning. The statistical evidence for this conclusion is shown to depend upon assumptions about (1) the asymptotic level reached and (2) the relative magnitude of the variance between conditions as compared to the variance of Ss within conditions. Evidence concerning these assumptions is reviewed, and it is argued that there was insufficient evidence to support a conclusion of negative transfer in part-whole free recall.

Discovering latent strategies

Abstract: Strategy mining is a new area of research about discovering strategies in decision-making. In this paper, we formulate the strategy-mining problem as a clustering problem, called the latent-strategy problem. In a latent-strategy problem, a corpus of data instances is given, each of which is represented by a set of features and a decision label. The inherent dependency of the decision label on the features is governed by a latent strategy. The objective is to find clusters, each of which contains data instances governed by the same strategy. Existing clustering algorithms are inappropriate to cluster dependency because they either assume feature independency (e.g., K-means) or only consider the co-occurrence of features without explicitly modeling the special dependency of the decision label on other features (e.g., Latent Dirichlet Allocation (LDA)). In this paper, we present a baseline unsupervised learning algorithm for dependency clustering. Our model-based clustering algorithm iterates between an assignment step and a minimization step to learn a mixture of decision tree models that represent latent strategies. Similar to the Expectation Maximization algorithm, our algorithm is grounded in the statistical learning theory. Different from other clustering algorithms, our algorithm is irrelevant-feature resistant and its learned clusters (modeled by decision trees) are strongly interpretable and predictive. We systematically evaluate our algorithm using a common law dataset comprised of actual cases. Experimental results show that our algorithm significantly outperforms K-means and LDA on clustering dependency.