Results 1  10
of
15
An Active Testing Model for Tracking Roads in Satellite Images
 IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
, 1995
"... We present a new approach for tracking roads from satellite images, and thereby illustrate a general computational strategy ("active testing") for tracking 1D structures and other recognition tasks in computer vision. Our approach is related to recent work in active vision on "where ..."
Abstract

Cited by 167 (5 self)
 Add to MetaCart
We present a new approach for tracking roads from satellite images, and thereby illustrate a general computational strategy ("active testing") for tracking 1D structures and other recognition tasks in computer vision. Our approach is related to recent work in active vision on "where to look next" and motivated by the "divideandconquer" strategy of parlor games such as "Twenty Questions." We choose "tests" (matched filters for short road segments) one at a time in order to remove as much uncertainty as possible about the "true hypothesis" (road position) given the results of the previous tests. The tests are chosen online based on a statistical model for the joint distribution of tests and hypotheses. The problem of minimizing uncertainty (measured by entropy) is formulated in simple and explicit analytical terms. To execute this entropy testing rule we then alternate between data collection and optimization: at each iteration new image data are examined and a new entropy minimizat...
Automatic Construction of Decision Trees from Data: A MultiDisciplinary Survey
 Data Mining and Knowledge Discovery
, 1997
"... Decision trees have proved to be valuable tools for the description, classification and generalization of data. Work on constructing decision trees from data exists in multiple disciplines such as statistics, pattern recognition, decision theory, signal processing, machine learning and artificial ne ..."
Abstract

Cited by 164 (1 self)
 Add to MetaCart
Decision trees have proved to be valuable tools for the description, classification and generalization of data. Work on constructing decision trees from data exists in multiple disciplines such as statistics, pattern recognition, decision theory, signal processing, machine learning and artificial neural networks. Researchers in these disciplines, sometimes working on quite different problems, identified similar issues and heuristics for decision tree construction. This paper surveys existing work on decision tree construction, attempting to identify the important issues involved, directions the work has taken and the current state of the art. Keywords: classification, treestructured classifiers, data compaction 1. Introduction Advances in data collection methods, storage and processing technology are providing a unique challenge and opportunity for automated data exploration techniques. Enormous amounts of data are being collected daily from major scientific projects e.g., Human Genome...
The Power of Decision Tables
 Proceedings of the European Conference on Machine Learning
, 1995
"... . We evaluate the power of decision tables as a hypothesis space for supervised learning algorithms. Decision tables are one of the simplest hypothesis spaces possible, and usually they are easy to understand. Experimental results show that on artificial and realworld domains containing only discre ..."
Abstract

Cited by 114 (5 self)
 Add to MetaCart
. We evaluate the power of decision tables as a hypothesis space for supervised learning algorithms. Decision tables are one of the simplest hypothesis spaces possible, and usually they are easy to understand. Experimental results show that on artificial and realworld domains containing only discrete features, IDTM, an algorithm inducing decision tables, can sometimes outperform stateoftheart algorithms such as C4.5. Surprisingly, performance is quite good on some datasets with continuous features, indicating that many datasets used in machine learning either do not require these features, or that these features have few values. We also describe an incremental method for performing crossvalidation that is applicable to incremental learning algorithms including IDTM. Using incremental crossvalidation, it is possible to crossvalidate a given dataset and IDTM in time that is linear in the number of instances, the number of features, and the number of label values. The time for incre...
Wrappers For Performance Enhancement And Oblivious Decision Graphs
, 1995
"... In this doctoral dissertation, we study three basic problems in machine learning and two new hypothesis spaces with corresponding learning algorithms. The problems we investigate are: accuracy estimation, feature subset selection, and parameter tuning. The latter two problems are related and are stu ..."
Abstract

Cited by 111 (7 self)
 Add to MetaCart
In this doctoral dissertation, we study three basic problems in machine learning and two new hypothesis spaces with corresponding learning algorithms. The problems we investigate are: accuracy estimation, feature subset selection, and parameter tuning. The latter two problems are related and are studied under the wrapper approach. The hypothesis spaces we investigate are: decision tables with a default majority rule (DTMs) and oblivious readonce decision graphs (OODGs).
Lookahead and Pathology in Decision Tree Induction
 Proceedings of the 14th International Joint Conference on Artificial Intelligence
, 1995
"... The standard approach to decision tree induction is a topdown, greedy algorithm that makes locally optimal, irrevocable decisions at each node of a tree. In this paper, we study an alternative approach, in which the algorithms use limited lookahead to decide what test to use at a node. We systemati ..."
Abstract

Cited by 51 (2 self)
 Add to MetaCart
The standard approach to decision tree induction is a topdown, greedy algorithm that makes locally optimal, irrevocable decisions at each node of a tree. In this paper, we study an alternative approach, in which the algorithms use limited lookahead to decide what test to use at a node. We systematically compare, using a very large number of decision trees, the quality of decision trees induced by the greedy approach to that of trees induced using lookahead. The main results of our experiments are: (i) the greedy approach produces trees that are just as accurate as trees produced with the much more expensive lookahead step; and (ii) decision tree induction exhibits pathology, in the sense that lookahead can produce trees that are both larger and less accurate than trees produced without it. 1. Introduction The standard algorithm for constructing decision trees from a set of examples is greedy induction  a tree is induced topdown with locally optimal choices made at each node, with...
Shape Recognition and Twenty Questions
 IN PROC. RECONNAISSANCE DES FORMES ET INTELLIGENCE ARTIFICIELLE (RFIA
, 1993
"... We formulate shape recognition as a coding problem. There is a finite list of possible "hypotheses"  shape classes and/or spatial positionings  and we wish to determine which one is true based on the results of various "tests," which are local image features. We use a decision ..."
Abstract

Cited by 6 (2 self)
 Add to MetaCart
We formulate shape recognition as a coding problem. There is a finite list of possible "hypotheses"  shape classes and/or spatial positionings  and we wish to determine which one is true based on the results of various "tests," which are local image features. We use a decision tree: each interior node is assigned one of the tests and each terminal node is assigned one of the hypotheses. The assignment of tests, or "strategy," is recursive: along each branch choose the next test to remove as much uncertainty as possible (as measured by entropy) about the true hypothesis. In contrast to the standard approach of "hypothesize and test," there is no repeated elicitation of hypotheses; instead, the "indexing" is dynamic and stochastic. We gradually formulate specific conjectures as the evolving distribution on hypotheses becomes increasingly peaked. We apply this "twenty questions" approach to the recognition of two types of linear, deformable structures: handwritten numerals and roads i...
EntropyBased Optimum Test Points Selection for Analog Fault Dictionary Techniques
 IEEE Transactions On Instrumentation And Measurement
, 2004
"... Abstract—An efficient method to select an optimum set of test points for dictionary techniques in analog fault diagnosis is proposed. This is done by searching for the minimum of the entropy index based on the available test points. First, the twodimensional integercoded dictionary is constructed ..."
Abstract

Cited by 4 (1 self)
 Add to MetaCart
Abstract—An efficient method to select an optimum set of test points for dictionary techniques in analog fault diagnosis is proposed. This is done by searching for the minimum of the entropy index based on the available test points. First, the twodimensional integercoded dictionary is constructed whose entries are measurements associated with faults and test points. The problem of optimum test points selection is, thus, transformed to the selection of the columns that isolate the rows of the dictionary. Then, the likelihood for a column to be chosen based on the size of its ambiguity set is evaluated using the minimum entropy index of test points. Finally, the test point with the minimum entropy index is selected to construct the optimum set of test points. The proposed entropybased method to select a local minimum set of test points is polynomial bounded in computational cost. The comparison between the proposed method and other reported test points selection methods is carried out by statistical experiments. The results indicate that the proposed method more efficiently and more accurately finds the locally optimum set of test points and is practical for large scale analog systems. Index Terms—Analog fault diagnosis, fault dictionary, rough set, test point. I.
Randomized Inquiries About Shape; an Application to Handwritten Digit Recognition
, 1994
"... We describe an approach to shape recognition based on asking relational questions about the arrangement of landmarks, basically localized and oriented boundary segments. The questions are grouped into highly structured inquiries in the form of a tree. There are, in fact, many trees, each constructed ..."
Abstract

Cited by 4 (1 self)
 Add to MetaCart
We describe an approach to shape recognition based on asking relational questions about the arrangement of landmarks, basically localized and oriented boundary segments. The questions are grouped into highly structured inquiries in the form of a tree. There are, in fact, many trees, each constructed from training data based on entropy reduction. The outcome of each tree is not a classification but rather a distribution over shape classes. The final classification is based on an aggregate distribution. The framework is nonEuclidean and there is no feature vector in the standard sense. Instead, the representation of the image data is graphical and each question is associated with a labeled subgraph. The ordering of the questions is highly constrained in order to maintain computational feasibility, and dependence among the trees is reduced by randomly subsampling from the available pool of questions. Experiments are reported on the recognition of handwritten digits. Although the amount ...
An Experimental Evaluation of Simplicity in Rule Learning
"... While recent research on rule learning has focussed largely on finding highly accurate hypotheses, we evaluate the degree to which these hypotheses are also simple, that is small. To realize this, we compare wellknown rule learners, such as CN2, RIPPER, PART, FOIL and C5.0 rules, with the benchmark ..."
Abstract
 Add to MetaCart
While recent research on rule learning has focussed largely on finding highly accurate hypotheses, we evaluate the degree to which these hypotheses are also simple, that is small. To realize this, we compare wellknown rule learners, such as CN2, RIPPER, PART, FOIL and C5.0 rules, with the benchmark system SL2 that explicitly aims at computing small rule sets with few literals. The results show that it is possible to obtain a similar level of accuracy as stateoftheart rule learners using much smaller rule sets.