Topic
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.
Papers published on a yearly basis
Papers
More filters
12 Nov 2012
TL;DR: This work exploits concepts from the "Transfer Learning" and "Imbalanced Learning" domains to expand the training set and permit standard models to be applied.
Abstract: The exponential growth of data dimensions presents an obstacle in informatics as data miners try to construct ever greater training sets to overcome the theoretical limitations of statistical learning theory. Machine learning models require a minimum set of samples within each label to develop a representative hypothesis. To overcome these bounds, we developed an algorithm that can extract samples from an auxiliary domain to augment the training set. Our work exploits concepts from the "Transfer Learning" and "Imbalanced Learning" domains to expand the training set and permit standard models to be applied. We present theoretical verification of our method and demonstrate the effectiveness of our framework with experimental results on real-world data.
1 citations
01 Jan 2005
TL;DR: It is proved that the standard method consisting in choosing a maximal program size depending on the number of examples might still result in programs of infinitely increasing size with their accuracy; a fitness biased by parsimony pressure is proposed.
Abstract: Universal Consistency, the convergence to the minimum possible error rate in learning through genetic programming (GP), and Code bloat, the excessive increase of code size, are important issues in GP. This paper proposes a theoretical analysis of universal consistency and code bloat in the framework of symbolic regression in GP, from the viewpoint of Statistical Learning Theory, a well grounded mathematical toolbox for Machine Learning. Two kinds of bloat must be distinguished in that context, depending whether the target function has finite description length or not. Then, the Vapnik-Chervonenkis dimension of programs is computed, and we prove that a parsimonious fitness ensures Universal Consistency (i.e. the fact that the solution minimizing the empirical error does converge to the best possible error when the number of examples goes to infinity). However, it is proved that the standard method consisting in choosing a maximal program size depending on the number of examples might still result in programs of infinitely increasing size with their accuracy; a fitness biased by parsimony pressure is proposed. This fitness avoids unnecessary bloat while nevertheless preserving the Universal Consistency.
1 citations
Journal Article•
TL;DR: Experimental results show that the LIB-SVM system can projects an accurate text classification, the method is practical and effective.
Abstract: SVM is a new machine learning method based on statistical learning theory, can achieve very good results in text classification?An accurate classification of the scientific project text contributes to scientific management and supervision of scientific project A automatic classification system of projects text based LIB-SVM is designed and implemented, and experimental results show that the system can projects an accurate text classification, the method is practical and effective
1 citations
Journal Article•
TL;DR: A new approach to Short Term Load Forecasting (STLF) using probabilistic neural network (PNN) is proposed in this work which results in a self organizing adaptive forecast model which adapts with changing load characteristics.
Abstract: A new approach to Short Term Load Forecasting (STLF) using probabilistic neural network (PNN) is proposed in this work. The present approach is different from other popular feed forward ANN methods as PNN uses statistical learning theory for learning from examples. The PNN processes the training data only once and therefore, it is very fast compared to back propagation ANN. In addition, PNN parameters and input variables selection is optimized using simulated annealing. The proposed method results in a self organizing adaptive forecast model which adapts with changing load characteristics. Since, the input selection and parameter identification is integrated with the forecaster, the approach becomes utility independent unlike other ANN approaches which are utility dependent. The comparison of present approach and its accuracy with other ANN and conventional approaches establish its superiority.
1 citations
12 Sep 2007
TL;DR: A proper theoretical framework, called reliable learning, for the analysis of consistency of learning techniques incorporating prior knowledge for the solution of pattern recognition problems is introduced by properly extending standard concepts of Statistical Learning Theory.
Abstract: A proper theoretical framework, called reliable learning, for the analysis of consistency of learning techniques incorporating prior knowledge for the solution of pattern recognition problems is introduced by properly extending standard concepts of Statistical Learning Theory. In particular, two different situations are considered: in the first one a reliable region is determined where the correct classification is known; in the second case the prior knowledge regards the correct classification of some points in the training set. In both situations sufficient conditions for ensuring the consistency of the Empirical Risk Minimization (ERM) criterion is established and an explicit bound for the generalization error is derived.
1 citations