An improved boosting algorithm and its implications on learning complexity

By Yoav Freund. 1992

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Optimally-Smooth Adaptive Boosting and Application to Agnostic Learning (2002, methodology): “Next we construct a “boosting tandem”, consisting ofAdaFlat “joined” with another boosting algorithm (this approach was first used by Freund (1992) and since then has become very popular for constructing boosting algorithms).”
An efficient membership-query algorithm for learning DNF with respect to the uniform distribution (1994, methodology): “The other tool we use is hypothesis boosting [44, 21, 22, 24].”
Selection of Relevant Features and Examples in Machine Learning (1997, methodology): “In the PAC model, learning algorithms need to roughly double the number of examples seen in order to halve their error rate (Schapire, 1990; Freund, 1992; Blumer et al., 1989).”
Boosting and Hard-Core Set Construction (2003, methodology): “ Freund (1992) describes a two-level boosting algorithm which works as follows: algorithm F1 is used to boost from accuracy (1/2 − γ ) to accuracy 1/4, and BOOSTING AND HARD-CORE SET CONSTRUCTION 227”
Text Classification Algorithms: A Survey (2019, methodology): “ting The boosting algorithm was ﬁrst introduced by R.E. Schapire [111] in 1990 as a technique for boosting the performance of a weak learning algorithm. This technique was further developed by Freund [112,113]. Figure9shows how a boosting algorithm works for 2D data sets, as shown we have labeled th”
On Agnostic Learning of Parities, Monomials, and Halfspaces (2009, methodology): “p-smooth). In Jackson’s result [Jac97], Freund’s boost-by-majority algorithm [Fre90] is used to produce distribution functions bounded by O( (2+ )) (for arbitrarily small constant ). More recently, Klivans and Servedio have observed [KS03] that a later algorithm by Freund [ Fre92 ] produces distribu”
Toward efficient agnostic learning (2004, methodology): “The result then follows from the "boosting" techniques of Schapire (1990) or Freund (1990; 1992) for converting a weak learning algorithm into a strong learning algorithm.”
Toward Efficient Agnostic Learning (1992, methodology): “Proof: The idea is that since T weakly approximates T, whenever the target function is from T, opt(T) will be significantly smaller than 1/2, and the agnostic learning algorithm effectively functions as a weak learning algorithm for T. The result then follows from the "boosting" techniques of Schapi”
Extracting comprehensible models from trained neural networks (1996, methodology): “Several such boosting algorithms have been developed (Schapire, 1990; Freund, 1990; Freund, 1992), and one called AdaBoost (Freund & Schapire, 1995) is the basis of the BBP algorithm.”