Accession Number : ADA616936


Title :   Machine Learning with Distances


Descriptive Note : Final rept. 28 Mar 2013-27 Mar 2015


Corporate Author : TOKYO INST OF TECHNOLOGY (JAPAN)


Personal Author(s) : Sugiyama, Masashi


Full Text : http://www.dtic.mil/dtic/tr/fulltext/u2/a616936.pdf


Report Date : 16 Feb 2015


Pagination or Media Count : 114


Abstract : Various machine learning tasks such as learning under non-stationarity, change detection, and dimensionality reduction can be solved by estimating some distance/ratio between two probability distributions. This project developed accurate and computationally efficient methods for estimating the distance/ratio from data, and demonstrated their usefulness in experiments. The principle idea is that when solving a problem of interest, we should not solve a more general sub-problem as an intermediate step, i,e., directly estimate the difference/ratio of the two distributions rather than estimating both separately and take the difference/ratio later. Types of the problems actually solved are change detection in time series, salient object detection in an image, measuring statistical independence, detection of structure change, covariance shift, class balance change, information maximization clustering.


Descriptors :   *LEARNING MACHINES , CLUSTERING , COVARIANCE , DATA MINING , EXPERIMENTAL DATA , GAUSSIAN NOISE , HEURISTIC METHODS , INFORMATION THEORY , KNOWLEDGE MANAGEMENT , LEAST SQUARES METHOD , OPTIMIZATION , PATTERN RECOGNITION , PERFORMANCE(ENGINEERING) , PROBABILITY DISTRIBUTION FUNCTIONS , STATISTICAL INFERENCE , STOCHASTIC PROCESSES , TIME SERIES ANALYSIS


Subject Categories : Statistics and Probability
      Cybernetics


Distribution Statement : APPROVED FOR PUBLIC RELEASE