Extracting Tree-Structured Representations of Trained Networks
نویسندگان
چکیده
A signiicant limitation of neural networks is that the representations they learn are usually incomprehensible to humans. We present a novel algorithm, Trepan, for extracting comprehensible, symbolic representations from trained neural networks. Our algorithm uses queries to induce a decision tree that approximates the concept represented by a given network. Our experiments demonstrate that Trepan is able to produce decision trees that maintain a high level of delity to their respective networks while being com-prehensible and accurate. Unlike previous work in this area, our algorithm is general in its applicability and scales well to large networks and problems with high-dimensional input spaces.
منابع مشابه
Extracting Thee-Structured Representations of Thained Networks
A significant limitation of neural networks is that the representations they learn are usually incomprehensible to humans. We present a novel algorithm, TREPAN, for extracting comprehensible , symbolic representations from trained neural networks. Our algorithm uses queries to induce a decision tree that approximates the concept represented by a given network. Our experiments demonstrate that T...
متن کاملUnderstanding Time-Series Networks: A Case Study in Rule Extraction
A significant limitation of neural networks is that the representations they learn are usually incomprehensible to humans. We have developed an algorithm, called TREPAN, for extracting comprehensible, symbolic representations from trained neural networks. Given a trained network, TREPAN produces a decision tree that approximates the concept represented by the network. In this article, we discus...
متن کاملConverting A Trained Neural Network To a Decision Tree DecText - Decision Tree Extractor
Neural Networks are very successful in acquiring hidden knowledge in datasets. Their most important weakness is that the knowledge they acquire is represented in a form not understandable to humans. Understandability problem of Neural Networks can be solved by extracting Decision Rules or Decision Trees from the trained network. There are several Decision Rule extraction methods and Mark Craven...
متن کاملBiological data mining with neural networks: implementation and application of a flexible decision tree extraction algorithm to genomic problem domains
In the past, neural networks have been viewed as classi5cation and regression systems whose internal representations were extremely di6cult to interpret. It is now becoming apparent that algorithms can be designed which extract understandable representations from trained neural networks, enabling them to be used for data mining, i.e. the discovery and explanation of previously unknown relations...
متن کاملStructured Attention Networks
Attention networks have proven to be an effective approach for embedding categorical inference within a deep neural network. However, for many tasks we may want to model richer structural dependencies without abandoning end-to-end training. In this work, we experiment with incorporating richer structural distributions, encoded using graphical models, within deep networks. We show that these str...
متن کامل