AdaBoost Parallelization on PC Clusters with Virtual Shared Memory for Fast Feature Selection

Galtier, Virginie; Pietquin, Olivier; Vialle, Stéphane

doi:10.1109/icspc.2007.4728281

Cited by 5 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently works have proposed parallel versions of Adaboost for classification problems, either by modifying the algorithm structure [16,18] to make it compatible with a parallel framework or using the original Adaboost in parallel hardware environments [10]. The method proposed in [18] works in two stages.…”

Section: Introductionmentioning

confidence: 99%

“…Parallel implementations of machine learning algorithms have gained importance [5,10,18,25] because dataset sizes appear to be growing considerably faster than computational capabilities. Recently works have proposed parallel versions of Adaboost for classification problems, either by modifying the algorithm structure [16,18] to make it compatible with a parallel framework or using the original Adaboost in parallel hardware environments [10].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Parallel Approach for Ensemble Learning with Locally Coupled Neural Networks

Valle

Saravia

Allende

et al. 2010

Neural Process Lett

View full text Add to dashboard Cite

Ensemble learning has gained considerable attention in different tasks including regression, classification and clustering. Adaboost and Bagging are two popular approaches used to train these models. The former provides accurate estimations in regression settings but is computationally expensive because of its inherently sequential structure, while the latter is less accurate but highly efficient. One of the drawbacks of the ensemble algorithms is the high computational cost of the training stage. To address this issue, we propose a parallel implementation of the Resampling Local Negative Correlation (RLNC) algorithm for training a neural network ensemble in order to acquire a competitive accuracy like that of Adaboost and an efficiency comparable to that of Bagging. We test our approach on both synthetic and real datasets from the UCI and Statlib repositories for the regression task. In particular, our fine-grained parallel approach allows us to achieve a satisfactory balance between accuracy and parallel efficiency.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Parallel Approach for Ensemble Learning with Locally Coupled Neural Networks

Valle

Saravia

Allende

et al. 2010

Neural Process Lett

View full text Add to dashboard Cite

show abstract

“…Also, many parallel machine algorithms for classification problems have been proposed [17,18,19,20,21,22,23]. However parallel algorithms are usually analyzed using metrics such as speedup or efficiency.…”

Section: Motivationmentioning

confidence: 99%

“…The additional element x 0 in x is always set to 1 and is needed to compute the additional element β 0 in β which is the bias factor, which is analogous to the y-intercept in a linear regression model. Thus Equation 1.17 can be rewritten as 19) and the problem of learning the model is reduced to picking the values of the vector β based on the values in the training set such that p is as close to the actual Y value for all samples in the training set.…”

Section: Logistic Regressionmentioning

confidence: 99%