CONVERGENCE PROPERTIES OF A LEARNING ALGORITHM, (2005)
In the learning process described by the algorithm, observations are made on individuals one at a time and the current estimate of the required partitioning may be adjusted after each observation, on...
Population theory for boosting ensembles (2004)
Tree ensembles are looked at in distribution space, that is, the limit case of "infinite" sample size. It is shown that the simplest kind of trees is complete in D-dimensional $L_2(P)$ space if the...
To dispel some of the mystery about what makes tree ensembles work, they are looked at in distribution space i.e. the limit case of "infinite" sample size. It is shown that the simplest kind of trees...
Synthetic Aperture Radar Signals: Formulations and Approaches for Data Analysis. (2002)
Lucero,Antonio B., Swerling,Peter, Breiman,Leo
This report discusses principles of synthetic aperture radar, properties of radar targets, characteristics of radar imagery, statistical analysis of radar imagery, and the application of modern data...
Topics in the Analysis and Optimization of Complex Systems. (2002)
Meisel,William S., Breiman,Leo
This document is a report of the results of research under AFOSR Contract No. F44620-71-C-0093 through Dec. 1975. The general thrust of the research is the development of approaches to improving the...
Topics in the Analysis and Optimization of Complex Systems. (2002)
Meisel,William S., Breiman,Leo
This one-year contract was in essence a continuation of a previous five year effort. The purpose was to discover more effective methods of analyzing high dimensional data sets. The motivation was the...
New Methods for Estimating Tail Probabilities and Extreme Value Disributions. (2002)
Breiman,Leo, Stone,Charles J., Gins,John D.
This research has focused on the problem of estimating probabilities in the upper tail of an underlying distribution and the corresponding quantiles based on a random sample from the distribution....
The limiting behavior of sums of functions of nearest neighbor distances is studied for an m dimensional sample. A central limit theorem and moment bounds for such sums, and an invariance principle...
Estimating Optimal Transformations for Multiple Regression and Correlation. (2002)
Breiman,Leo, Friedman,Jerome H.
Nonlinear transformation of variables is a commonly used practice in regression problems. Two common goals are stabilization of error variance and asymmetrization/normalization of error distribution....
Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The...
Random Forests: Finding Quasars (2001)
Leo Breiman, Michael Last, John Rice
this paper, we discuss an example in which we classify objects as quasars or non-quasars using the combined results of a radio survey and an optical survey. Such classi cation helps guide the choice...
Randomizing Outputs To Increase Prediction Accuracy (2001)
Introduction In recent research in combining predictors, it has been recognized that the critical thing to success in combining low-bias predictors such as trees and neural nets has been through...
Statistical Modeling: The Two Cultures (with comments and a rejoinder by the author) (2001)
There are two cultures in the use of statistical modeling to reach conclusions from data. One assumes that the data are generated by a given stochastic data model. The other uses algorithmic models...
Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The...
Bagging predictors is a method for generating multiple versions of a predictor and using these to get an aggregated predictor. The aggregation averages over the versions when predicting a numerical...
Pasting Bites Together For Prediction In Large Data Sets And On-Line (2000)
The size of many data bases have grown to the point where they cannot fit into the fast memory of even large memory machines, to say nothing of current workstations. If what we want to do is to use...
Halfhalf Bagging And Hard Boundary Points (2000)
Introduction Half&half bagging is a method for producing combinations of classifiers having low generalization error. The basic idea is straightforward and intuitive--suppose k classifiers have been...
Prediction Games and Arcing Algorithms (2000)
The theory behind the success of adaptive reweighting and combining algorithms (arcing) such as Adaboost (Freund and Schapire [1995].[1996]) and others in reducing generalization error has not been...
Bias, Variance , And Arcing Classifiers (2000)
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. To study this, the concepts of bias and variance of a...
Some Infinite Theory for Predictor Ensembles (2000)
To dispel some of the mystery about what makes tree ensembles work, they are looked at in distribution space i.e. the limit case of "infinite" sample size. It is shown that the simplest kind of trees...
Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The...
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. One of the more effective is bagging (Breiman [1996a] )...
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. One of the more effective is bagging (Breiman [1996a] )...
Prediction Games And Arcing Algorithms (1998)
The theory behind the success of adaptive reweighting and combining algorithms (arcing) such as Adaboost (Freund and Schapire [1995, 1996a]) and others in reducing generalization error has not been...
Topics in the Analysis and Optimization of Complex Systems. (1998)
Meisel,William S., Collins,David C., Breiman,Leo
The general thrust of the research is the development of approaches to improving the utility of computer-oriented quantitative methods in practical decision-making and system design. The research was...
Halfhalf Bagging And Hard Boundary Points (1998)
Introduction Half&half bagging is a method for producing combinations of classifiers having low generalization error. The basic idea is straightforward and intuitive--suppose k classifiers have been...
Arcing classifier (with discussion and a rejoinder by the author) (1998)
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. One of the more effective is bagging. Here, modified...
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. One of the more effective is bagging (Breiman [1996a] )...
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. One of the more effective is bagging (Breiman [1996a] )...
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. One of the more effective is bagging (Breiman [1996a] )...
Recent work has shown that adaptively reweighting the training set, growing a classifier using the new weights, and combining the classifiers constructed to date can significantly decrease...
Bagging predictors is a method for generating multiple versions of a predictor and using these to get an aggregated predictor. The aggregation averages over the versions when predicting a numerical...
Bias, Variance, and Arcing Classifiers (1997)
Recent work has shown that combining multiple versions of unstable classifiers such as trees or neural nets results in reduced test set error. To study this, the concepts of bias and variance of a...
Heuristics of instability and stabilization in model selection (1996)
In model selection, usually a "best" predictor is chosen from a collection ${\hat{\mu}(\cdot, s)}$ of predictors where $\hat{\mu}(\cdot, s)$ is the minimum least-squares predictor in a collection...
Pasting Bites Together For Prediction In Large Data Sets And On-Line (1996)
The size of many data bases have grown to the point where they cannot fit into the fast memory of even large memory machines, to say nothing of current workstations. If what we want to do is to use...
In bagging, predictors are constructed using bootstrap samples from the training set and then aggregated to form a bagged predictor. Each bootstrap sample leaves out about 37% of the examples. These...
Distribution Based Trees Are More Accurate (1996)
Classification trees are attractive in that they present a simple and easily understandable structure. But on many data sets their accuracy is far from optimal. Much of this lack of accuracy is due...
Bagging predictors is a method for generating multiple versions of a predictor and using these to get an aggregated predictor. The aggregation averages over the versions when predicting a numerical...
Homogeneous processes--[microform] /--by Leo Breiman. (1954)
Thesis (Ph. D. in Mathematics)--University of California, Berkeley, June 1954.
Homogeneous processes /--by Leo Breiman. (1954)
Thesis (Ph. D. in Mathematics)--University of California, Berkeley, June 1954.
Two problems in the dynamics of cavitation bubbles--[microform] /--by Leo Breiman. (1950)
Thesis (M.A.)--Columbia University, 1950.