The Cruncher: Automatic Concept Formation Using Minimum Description Length
نویسندگان
چکیده
We present The Cruncher, a simple representation framework and algorithm based on minimum description length for automatically forming an ontology of concepts from attribute-value data sets. Although unsupervised, when The Cruncher is applied to an animal data set, it produces a nearly zoologically accurate categorization. We demonstrate The Cruncher’s utility for finding useful macro-actions in Reinforcement Learning, and for learning models from uninterpreted sensor data. We discuss advantages The Cruncher has over concept lattices and hierarchical clustering.
منابع مشابه
Automatic Reconnection of Linear Segments by the Minimum Description Length Principle
The automatic reconnection of linear segments is a problem often encountered in image analysis. This article proposes a procedure for performing this task. Tuning parameters of the proposed procedure can either be chosen manually, or chosen automatically by a method developed in this note. This automatic method is based on the minimum description length principle. The procedure is applied to so...
متن کاملConcept formation in noisy domains
In inductive logic programming, concept formation in noisy domains can be considered as learning from noisy data. This paper describes an approach to learning from noisy examples with an approximate theory. Although this kind of learning is more close to practical application, there are few systems can deal with such problems. The proposed learning approach includes a theory preference criterio...
متن کاملSegmenting by Compression Using Linear Scale-Space and Watersheds
Automatic segmentation is performed using watersheds of the gradient magnitude and compression techniques. Linear Scale-Space is used to discover the neighbourhood structure and catchment basins are locally merged with Minimum Description Length. The algorithm can form a basis for a large range of automatic segmentation algorithms based on watersheds, scale-spaces, and compression.
متن کاملAttribute Value Selection Considering the Minimum Description Length Approach and Feature Granularity
In this paper we introduce a new approach to automatic attribute and granularity selection for building optimum regression trees. The method is based on the minimum description length principle (MDL) and aspects of granular computing. The approach is verified by giving an example using a data set which is extracted and preprocessed from an operational information system of the Components Toolsh...
متن کاملABN: A Fast, Greedy Bayesian Network Classifier
Adaptive Bayes Network (ABN) is a fast algorithm for constructing Bayesian Network classifiers using Minimum Description Length (MDL) and automatic feature selection. ABN does well in domains where Naive Bayes fares poorly, and in other domains is, within statistical bounds, at least as good a classifier.
متن کامل