
	Where results make sense

Topic: AdaBoost

In the News (Sun 15 Nov 09)

Northern Trust

Marvin Harrison

Peter Chernin

Gary Locke

Match Play

Penelope Cruz

Mickey Rourke

AR Rahman

Danny Boyle

Hugh Jackman

	AdaBoost - Wikipedia, the free encyclopedia
		AdaBoost, short for Adaptive Boosting, was formulated by Yoav Freund and Robert Schapire.
		AdaBoost is adaptive in the sense that subsequent classifiers built are tweaked in favor of those instances misclassified by previous classifiers.
		AdaBoost is sensitive to noisy data and outliers.
	en.wikipedia.org /wiki/AdaBoost (224 words)

	Caltech CNSE - Shenoy - full report
		AdaBoost is probably the most popular algorithm among the boosting family which generates a linear combination of weak hypotheses.
		AdaBoost can be viewed as a special case of AnyBoost, a general gradient descent in a function space.
		One explanation to this is that AdaBoost improves the margins of the training examples even after all the examples have positive margins, and larger margins imply better out-of-sample performance.
	www.erc.caltech.edu /Research/reports/li-full.html (709 words)

	NASA :: Intelligent Systems (Site not responding. Last check: 2007-10-24)
		AdaBoost is one of the best-known and highest-performing ensemble classifier learning algorithms, but its error rate increases dramatically with an increase in noise.
		When AdaBoost generates the next model, misclassified examples are given more attention, with the aim of correcting the mistakes of earlier models.
		As a result, AdaBoost is known to increase the weights of noisy examples excessively at the expense of most of the training data.
	ic-www.arc.nasa.gov /story.php?id=217&sec= (363 words)

	[No title]
		Adaboost is one of the most successful classification methods in use.
		AdaBoost is a classification technique based on a weighted combination of different realizations of a same base model.
		The procedure to regularize AdaBoost consists in shorting data points by hardness, as emerging from analysis of the evolution of AdaBoost weights, and in progressively eliminating the hardest from the data set.
	mpa.itc.it /papers/biblio04-abstracts.html (3663 words)

	Face Verification Using Gabor Wavelets and AdaBoost
		For the large example size, AdaBoost algorithm selected the top 20 features from the first client to the eighth client as shown in Fig 4.
		The results show that the selected features are randomly distributed in the face area rather than concentrated on some regions of the faces as the results from the small example size.
		AdaBoost selects the top 20 significant features which distinguish a specified client from other subjects in the face database.
	www.personal.rdg.ac.uk /~sir02mz/Research/AdaBoost.html (2477 words)

	[No title] (Site not responding. Last check: 2007-10-24)
		In order to study the convergence properties of the AdaBoost algorithm, we reduce AdaBoost to a nonlinear iterated map and study the evolution of its weight vectors.
		Using this unusual technique, we are able to show that AdaBoost does not always converge to a maximum margin combined classifier, answering an open question.
		In addition, we show that "non-optimal" AdaBoost (where the weak learning algorithm does not necessarily choose the best weak classifier at each iteration) may fail to converge to a maximum margin classifier, even if "optimal" AdaBoost produces a maximum margin.
	jmlr.csail.mit.edu /papers/v5/rudin04a.html (161 words)

	Paper: Using Boosting to Improve a Hybrid HMM/Neural Network Speech Recognizer :: Holger Schwenk (Site not responding. Last check: 2007-10-24)
		There is also recent evidence that AdaBoost may very well over t if we combine several hundred thousands classi ers [8] and [5] reports severe performance degradations of AdaBoost when adding 20% noise on the class-labels.
		In the third section we shown how AdaBoost can be applied to this task and we report results on the Numbers95 corpus and compare them with other classier combination techniques.
		ADABOOST AdaBoost, constructs a composite classier by sequentially training classi ers while putting more and more emphasis on certain patterns.
	computing.breinestorm.net /boosting+algorithm+learning+adaboost+improve (714 words)

	Soft Margins for AdaBoost (ResearchIndex)
		Abstract: Recently ensemble methods like AdaBoost were successfully applied to character recognition tasks, seemingly defying the problems of overfitting.
		This paper shows that although AdaBoost rarely overfits in the low noise regime it clearly does so for higher noise levels.
		Central for understanding this fact is the margin distribution and we find that AdaBoost achieves -- doing gradient descent in an error function with respect to the margin -- asymptotically a hard margin distribution, i.e.
	citeseer.ist.psu.edu /657521.html (535 words)

	Rudin abstract
		Statistical learning algorithms are wildly popular now due to their excellent performance on many types of data; one of the most successful learning algorithms is AdaBoost, which is a classification algorithm designed to construct a "strong" classifier from a "weak" learning algorithm.
		We simplify AdaBoost to reveal a nonlinear iterated map and analyze its behavior in specific cases.
		In this talk, I will introduce AdaBoost, describe our analysis of AdaBoost when viewed as a dynamical system, and briefly mention a new boosting algorithm which always maximizes the margin with a fast convergence rate.
	www.math.nyu.edu /~gunturk/Seminar_Data/Rudin_abstract.html (292 words)

	Soft Margins for AdaBoost - Ratsch, Onoda, Muller (ResearchIndex) (Site not responding. Last check: 2007-10-24)
		Abstract: Recently ensemble methods like AdaBoost have been applied successfully in many problems, while seemingly defying the problems of overfitting.
		AdaBoost rarely overfits in the low noise regime, however, we show that it clearly does so for higher noise levels.
		AdaBoost can be viewed as a constraint gradient descent in an error function with respect to the margin.
	citeseer.ist.psu.edu /264831.html (337 words)

	Discretizing Continuous Attributes in AdaBoost for Text Categorization
		Researchers from the University of Padova and from ISTI-CNR, Pisa, are undertaking a collaborative effort aimed at producing better best text classification strategies through the design of methods for the discretization of continuous attributes.
		In this work we make use of two algorithms, called AdaBoost.MH and AdaBoost.MH(KR), which are based on the notion of "adaptive boosting", a version of boosting in which members of the committee can be sequentially generated after learning from the classification mistakes of previously generated members of the same committee.
		AdaBoost.MH is a realization of the well-known AdaBoost algorithm, which is specifically aimed at multi-label TC (ie the TC task in which any number of categories may be assigned to each document), and which uses 'decision stumps' (ie decisions trees composed of a root and two leaves only) as weak hypotheses.
	www.ercim.org /publication/Ercim_News/enw52/sebastiani.html (698 words)

	Department of Computer Science
		In particular, we show that AdaBoost is a stability-preserving operation: if the "input" (the weak learner) to AdaBoost is stable, then the "output" (the strong learner) is almost-everywhere stable.
		Because classifier combination schemes such as AdaBoost have greatest effect when the weak learner is weak, we discuss weakness and its implications.
		We also show that the notion of almost-everywhere stability is sufficient for good bounds on generalization error.
	www.cs.uchicago.edu /research/publications/techreports/TR-2001-30 (211 words)

	TTIC Abstract (Site not responding. Last check: 2007-10-24)
		One of the most successful and popular learning algorithms is AdaBoost, which is a classification algorithm designed to construct a "strong" classifier from a "weak" learning algorithm.
		We then analyze the convergence of AdaBoost for cases where cyclic behavior is found; this cyclic behavior provides the key to answering the question of whether AdaBoost always maximizes the margin.
		In this talk, I will introduce AdaBoost, describe our analysis of AdaBoost when viewed as a dynamical system, briefly mention a new boosting algorithm which always maximizes the margin with a fast convergence rate, and if time permits, I will reveal a surprising new result about AdaBoost and the problem of bipartite ranking.
	ttic.uchicago.edu /events/event_detail.php?event_id=87 (247 words)

	Regularizing AdaBoost (ResearchIndex)
		Also for noisy data boosting will try to enforce a hard margin and thereby give too much weight to outliers, which then leads to the dilemma of non-smooth fits and overfitting.
		Therefore we propose three algorithms to allow for soft margin classification by introducing regularization with slack variables into the boosting concept: (1) AdaBoost reg and...
		9 An asymptotic analysis of adaboost in the binary classificat..
	citeseer.ist.psu.edu /69807.html (372 words)

	AdaBoost (Site not responding. Last check: 2007-10-24)
		AdaBoost is a boosting algorithm, running a given weak learner several times on slightly altered training data, and combining the hypotheses to one final hypothesis, in order to achieve higher accuracy than the weak learner's hypothesis would have.
		The main idea of AdaBoost is to assign each example of the given training set a weight.
		At the beginning all weights are equal, but in every round the weak learner returns a hypothesis, and the weights of all examples classified wrong by that hypothesis are increased.
	kiew.cs.uni-dortmund.de:8001 /mlnet/instances/81d91e8d-dc15ed23e9 (339 words)

	AdaBoost with Totally Corrective Updates for Fast Face Detection (Site not responding. Last check: 2007-10-24)
		An extension of the AdaBoost learning algorithm is proposed and brought to bear on the face detection problem.
		A cascaded face detector of the Viola-Jones type is built using AdaBoost with the Totally Corrective Update.
		The same detection and false positive rates are achieved with a detector that is 20% faster and consists of only a quarter of the weak classifiers needed for a classifier trained by standard AdaBoost.
	csdl2.computer.org /persagen/DLAbsToc.jsp?resourcePath=/dl/proceedings/&toc=comp/proceedings/fgr/2004/2122/00/2122toc.xml&DOI=10.1109/AFGR.2004.1301573 (250 words)

	Process consistency for AdaBoost, Wenxin Jiang
		Recent experiments and theoretical studies show that AdaBoost can overfit in the limit of large time.
		We show under general regularity conditions that during the process of AdaBoost a consistent prediction is generated, which has the prediction error approximating the optimal Bayes error as the sample size increases.
		This result suggests that, while running the algorithm forever can be suboptimal, it is reasonable to expect that some regularization method via truncation of the process may lead to a near-optimal performance for sufficiently large sample size.
	projecteuclid.org /getRecord?id=euclid.aos/1079120128 (366 words)

	Cynthia Rudin (Site not responding. Last check: 2007-10-24)
		In fact, it turns out that AdaBoost (which was not designed for ranking), is asymptotically just as good for ranking as RankBoost is. The longer version of this paper (co-authored with Robert Schapire) will be available soon.
		Also, there is an interesting result about AdaBoost in the longer version of the paper, namely, we derive a direct relationship between AdaBoost's edge values and its asymptotic margin.
		For instance, I have proved that AdaBoost may converge to a classifier with margin significantly below the maximum value; this was contrary to the widely believed conjecture that AdaBoost always converges to a maximum margin solution.
	www.cns.nyu.edu /~rudin/main.html (947 words)

	The interaction of stability and weakness in AdaBoost (ResearchIndex) (Site not responding. Last check: 2007-10-24)
		In particular, we show that AdaBoost is a stabilitypreserving operation: if the "input" (the weak learner) to AdaBoost is stable, then the "output" (the strong learner) is almost-everywhere stable.
		Because classifier combination schemes such as AdaBoost have greatest e#ect when the weak learner is weak, we discuss weakness and its implications.
		We also show that the notion of almost-everywhere stability is...
	citeseer.ist.psu.edu /kutin01interaction.html (307 words)

	Graphics & Media lab: Modest AdaBoost (Site not responding. Last check: 2007-10-24)
		During the research we have devised a new classification algorithm, based on weak classifier boosting approach.
		Our method, called "Modest AdaBoost", was implemented in MatLab enviroment and compared to well known Gentle AdaBoost scheme.
		Alexander Vezhnevets, Vladimir Vezhnevets "'Modest AdaBoost' - Teaching AdaBoost to Generalize Better".
	graphics.cs.msu.su /en/research/boosting (111 words)

	GVU/Research/Technical Reports/05-23
		In this paper, we provide a Bayesian perspective of boosting framework, which we refer to as Bayesian Integration.
		Through this perspective, we prove the standard ADABOOST is a special case of the naive
		Bayesian tree with a mapped conditional probability table and a particular weighting schema.
	www-static.cc.gatech.edu /gvu/research/tr/tr05_22.html (131 words)

	Boosting
		Boosting is a general method of producing a very accurate prediction rule by combining rough and moderately inaccurate "rules of thumb." Much recent work has been on the "AdaBoost" boosting algorithm and its extensions.
		Here is an overview of boosting focusing especially on AdaBoost:
		A discussion of ``Process consistency for AdaBoost'' by Wenxin Jiang, ``On the Bayes-risk consistency of regularized boosting methods'' by Gábor Lugosi and Nicolas Vayatis, ``Statistical behavior and consistency of classification methods based on convex risk minimization'' by Tong Zhang.
	www.cs.princeton.edu /~schapire/boost.html (801 words)

	DiSC - The application of AdaBoost for distributed, scalable and on-line learning (Site not responding. Last check: 2007-10-24)
		DiSC - The application of AdaBoost for distributed, scalable and on-line learning
		The application of AdaBoost for distributed, scalable and on-line learning
		Wei Fan, Salvatore J. Stolfo, and Junxin Zhang
	www.sigmod.org /sigmod/disc/p_theapplicationowesaj.htm (30 words)

Try your search on: Qwika (all wikis)