Active Learning for Regression Based on Query by Committee

Burbidge, Robert; Rowland, Jem J.; King, Ross D.

doi:10.1007/978-3-540-77226-2_22

Cited by 120 publications

(78 citation statements)

References 14 publications

(28 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…2) P-FV W : P-FV W is another variance reduction technique, which is to minimize the full variance expected over both training output noise and training examples [13]. 3) QBC: The QBC for regression approach chooses the data point, which has the largest variance among the members' prediction [23], [26]. 4) Greedy: The greedy approach aims to select the new example having the largest minimum distance from labeled data [17].…”

Section: B Comparison Methods and Evaluation Metricsmentioning

confidence: 99%

“…Cohn et al [14] presented an AL strategy with a similar motivation, which chooses the example, minimizing the output variance to reduce generalization error. Freund et al [23] suggested that the QBC framework could be applied to regression cases where the outputs are continuous, which is related to the variance-based QBC [26]. Yu and Kim [17] provided passive sampling heuristics based on the geometric characteristics of data.…”

Section: B Active Learning For Regressionmentioning

confidence: 99%

See 1 more Smart Citation

Batch Mode Active Learning for Regression With Expected Model Change

Cai

Zhang

2017

IEEE Trans. Neural Netw. Learning Syst.

View full text Add to dashboard Cite

While active learning (AL) has been widely studied for classification problems, limited efforts have been done on AL for regression. In this paper, we introduce a new AL framework for regression, expected model change maximization (EMCM), which aims at choosing the unlabeled data instances that result in the maximum change of the current model once labeled. The model change is quantified as the difference between the current model parameters and the updated parameters after the inclusion of the newly selected examples. In light of the stochastic gradient descent learning rule, we approximate the change as the gradient of the loss function with respect to each single candidate instance. Under the EMCM framework, we propose novel AL algorithms for the linear and nonlinear regression models. In addition, by simulating the behavior of the sequential AL policy when applied for k iterations, we further extend the algorithms to batch mode AL to simultaneously choose a set of k most informative instances at each query time. Extensive experimental results on both UCI and StatLib benchmark data sets have demonstrated that the proposed algorithms are highly effective and efficient.

show abstract

Section: B Comparison Methods and Evaluation Metricsmentioning

confidence: 99%

Section: B Active Learning For Regressionmentioning

confidence: 99%

Batch Mode Active Learning for Regression With Expected Model Change

Cai

Zhang

2017

IEEE Trans. Neural Netw. Learning Syst.

View full text Add to dashboard Cite

show abstract

“…Bu problemin çözümüne yönelik geliştirilen etkili bir yaklaşım aktif öğrenmedir. Uzaktan algılama alanında aktif öğrenme yöntemleri çoğunlukla sınıflandırma problemleri için geliştirilmiştir [3] ve regresyon problemleri için sınırlı sayıda araştırma bulunmaktadır [4]. ) hesapladıktan sonra en yüksek yoğunluğa sahip h p ≤ küme seçilmektedir.…”

Section: Gđrđşunclassified

Biophysical parameters estimation from remotely sensed images by a multiple criteria active learning method

Demir

Bruzzone

2014

2014 22nd Signal Processing and Communications Applications Conference (SIU)

View full text Add to dashboard Cite

Özetçe-Bu çalışmada uzaktan algılama görüntülerinden biyofiziksel parametrelerin kestirimi amacıyla ε-duyarsız destek vektör regresyon (DVR) yöntemi için çoklu ölçütlü özgün bir aktif öğrenme yöntemi önerilmektedir. Önerilen yöntem, en bilgi verici ve en temsilci örnekleri, örneklerin i) ilgililiği, ii) çeşitliliği ve iii) yoğunluğu olmak üzere üç farklı ölçütünü birlikte kullanarak seçmektedir. Bu ölçütler ε-duyarsız DVR özelliklerine göre tanımlanmaktadır ve özgün kümeleme temelli yaklaşımlar kullanılarak arda arda iki aşamada gerçekleştirilmektedir. Đlk aşamada DVR ε-tüpünün dışarısında yer alan etiketlenmemiş örnekler en ilgili örnekler olarak seçilmektedir. Đkinci aşamada ise en ilgili örnekler arasından çeşitliliği arttırmak amacıyla birbirinden en farklı örnekler DVR öğrenme fonksiyonunu daha iyi modelleyebilmek için öznitelik uzayının en yoğun örnekli bölgelerinden seçilmektedir. Önerilen yöntem ağaçların gövde hacminin kestirimine uygulanmıştır ve deneysel sonuçlar önerilen yöntemin başarımını göstermektedir. Anahtar Kelimeler-regresyon, parametre kestirimi, aktif öğrenme, destek vektör regresyonu. Abstract-This paper presents a novel multıple criteria active learning method developed in the framework of εinsensitive support vector regression (SVR) to estimate biophysical parameters from remotely sensed images. The proposed active learning method chooses the most informative and representative unlabeled samples by jointly evaluating three criteria: i) relevancy, ii) diversity, and iii) density of samples. All three criteria are implemented according to the SVR properties and are applied in two clustering-based successive steps. In the first step, a novel measure to select the most relevant samples that have high probability to be located either outside or on the boundary of the ε-tube of SVR is defined. In the second step, a novel measure to select diverse samples among the relevant patterns from the high density regions in the feature space is defined to better model the SVR learning function. Experiments applied to the estimation of tree stem volume show the effectiveness of the proposed method.

show abstract

“…Yu et al [43] proposed a transductive active learning approach which exploited the presence of unlabeled data in a linear regression learning problem. Burbidge et al [44] presented a variance based Query by…”

Section: Active Learning For Regressionmentioning

confidence: 99%

Batch Mode Active Learning for Multimedia Pattern Recognition

Chakraborty

Balasubramanian

Panchanathan

2012

2012 IEEE International Symposium on Multimedia

View full text Add to dashboard Cite

The rapid escalation of technology and the widespread emergence of modern technological equipments have resulted in the generation of humongous amounts of digital data (in the form of images, videos and text). This has expanded the possibility of solving real world problems using computational learning frameworks. However, while gathering a large amount of data is cheap and easy, annotating them with class labels is an expensive process in terms of time, labor and human expertise. This has paved the way for research in the field of active learning. Such algorithms automatically select the salient and exemplar instances from large quantities of unlabeled data and are effective in reducing human la-

show abstract

Active Learning for Regression Based on Query by Committee

Cited by 120 publications

References 14 publications

Batch Mode Active Learning for Regression With Expected Model Change

Batch Mode Active Learning for Regression With Expected Model Change

Biophysical parameters estimation from remotely sensed images by a multiple criteria active learning method

Batch Mode Active Learning for Multimedia Pattern Recognition

Contact Info

Product

Resources

About