Online Passive-Aggressive Active learning

Lü, Jing; Zhao, Peilin; Hoi, Steven C. H.

doi:10.1007/s10994-016-5555-y

Cited by 67 publications

(48 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These are all listed in the Supplement (Part II, Section 1). Specifically, we selected 10 predictor algorithms: 1) LOLIMOT (Local Linear Model Trees) [35,36], 2) RBF (Radial basis Function) [37], 3) MLP-BP (Multilayer Perceptron-Back propagation) [38,39], 4) LASSOLAR (Least Absolute Shrinkage and Selection Operator -Least Angle Regression) [40,41], 5) RFA (Random Forest Algorithm) [42,43], 6) RNN (Recurrent Neural Network) [44,45], 7) BRR (Bayesian Ridge Regression) [46][47][48], 8) DTC (Decision Tree Classification) [49][50][51], 9) PAR (Passive Aggressive Regression) [52][53][54], 10) Thiel-Sen Regression [55-57] and 11) ANFIS (Adaptive neuro fuzzy inference system) [58,59]. In this work, we automatically adjusted intrinsic parameters such as the number of neurons and number of layers in the predictor algorithms etc.…”

Section: Predictor Algorithms and Utilizing Automated Machine Learninmentioning

confidence: 99%

Machine learning methods for optimal prediction of motor outcome in Parkinson’s disease

et al. 2020

View full text Add to dashboard Cite

Section: Predictor Algorithms and Utilizing Automated Machine Learninmentioning

confidence: 99%

Machine learning methods for optimal prediction of motor outcome in Parkinson’s disease

et al. 2020

View full text Add to dashboard Cite

“…To address this problem, in this section, we develop online AL algorithms based CSS schemes to predict the channel availability. First, we exploit the margin based online AL algorithms such as CBGZ [12] and PAAL [13] and we adapt them to fit in the proposed CSS problem. Next, we propose a reduced threshold and SGD update based online AL algorithm for CSS.…”

Section: System Modelmentioning

confidence: 99%

“…If yes, the classifier acquires the true label from the system and follows the standard perceptron approach to update the classifier. Subsequently, the online passive aggressive AL (PAAL) algorithm [13] is considered for the proposed CSS problem. This algorithm follows the similar idea of CBGZ algorithm for querying, but utilizes an efficient PA learning strategy to update the classifier.…”

Section: Introductionmentioning

confidence: 99%

Margin-Based Active Online Learning Techniques for Cooperative Spectrum Sharing in CR Networks

Kumar

Lagunas

Sharma

et al. 2019

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

View full text Add to dashboard Cite

In this paper, we consider a problem of acquiring accurate spectrum availability information in the Cooperative Spectrum Sensing (CSS) based Cognitive Radio Networks (CRNs), where a fusion center collects the sensing information from all the sensing nodes within the network, analyzes the information and determines the spectrum availability. Although Machine Learning (ML) techniques have been recently applied to enhance the cooperative sensing performance in CRNs, they are mostly supervised learning based techniques and need a significant amount of labeled data, which is difficult to acquire in practice. Towards relaxing this requirement of large labeled data of supervised learning, we focus on Active Learning (AL), where the fusion center can query the label of the most uncertain cooperative sensing measurements. This is particularly relevant in CRN environments where primary user behavior changes in a quick manner. In this regard, we briefly review the existing AL techniques and adapt them to the considered CSS based CRNs. More importantly, we propose a novel margin based active on-line learning algorithm that selects the instance to be queried and updates the classifier by using the Stochastic Gradient Descent (SGD) technique. In this approach, whenever an unlabeled instance is presented, the proposed AL algorithm compares the margin of instance with a threshold to decide whether it should query a label or not. Supporting results based on numerical simulations show that the proposed method has significant advantages on classification and detection performances, and timecomplexity as compared to state-of-the-art techniques.

show abstract

“…Such a learning situation goes by the name of active learning. In pool-based and online environments [3,4], active learning has received widely attention and research.…”

Section: Introductionmentioning

confidence: 99%

“…Such a learning situation goes by the name of active learning. In pool-based and online environments [3,4], active learning has received widely attention and research.In the data stream setting, active learning is further divided into online active learning and active learning in data streams. The main difference between the two branches is whether the concept drifts exist or not.…”

mentioning

confidence: 99%

A Novel Sampling Strategy for Active Learning over Evolving Stream Data

Zhang¹,

Cao²,

Li³

et al. 2017

Proceedings of the 2nd International Conference on Computer Engineering, Information Science &Amp; Application Technology (ICCI

View full text Add to dashboard Cite

Abstract. In classification tasks, data labeling is an expensive and time-consuming process, hence, active learning which query labels for a small representative portion of data, is becoming increasingly important. However, few works consider the challenges from data steam setting because most of the active learning method is designed for non-streaming setting. Be based upon the status quo, after synthesizing the evidence-based uncertainty sampling strategy and split sampling strategy above, we propose a new sampling strategy for active learning over evolving stream data, which can take full advantages of the strengths of each. First, the original data stream is randomly divided into two substreams. Instances from one sub-stream are labeled according to the high evidence-focused uncertainty strategy, while instances from the other sub-stream are marked by the random strategy for detecting true concept drifts. Second, we introduce a sliding window in the high evidence-focused uncertainty strategy, finding out whether an instance is the conflict-uncertainty instance or not. Clearly, our strategy solves the issue of the effective use of evidence in data streams setting, and can choose more representative instances over evolving data streams for training a model. Finally, in experiments over four benchmark datasets, compared with state-of-art active learning strategies, the result illustrates good predictive performance of our proposed approach. IntroductionNowadays, more and more data are being generated continuously by networks, such as sensor networks, social networks, web applications and financial activities etc. Unlike traditional datasets, data items within a data stream are temporally ordered, fast-changing, generally large-scale, and potentially infinite [1].For learning predictive models on data stream, it is possible to access the true labels of the instances continuously. Unfortunately, inherent labeled instances in data streams are very scarce in practice. Conversely, a very limited number of labeled instances can be collected, and they can hardly provide enough information to train models with good generalization capabilities [2]. However, the manual labeling is expensive, especially in terms of time. Besides, in pace with time, the relationship between attributes and labels might change, such as spam identification and vaccine production. In order to know the true label, it is needed to scan the mail or make a laboratory test, which is timeconsuming. Hence, querying labels for a small representative subset of all stream data, has become an effective solution. Such a learning situation goes by the name of active learning. In pool-based and online environments [3,4], active learning has received widely attention and research.In the data stream setting, active learning is further divided into online active learning and active learning in data streams. The main difference between the two branches is whether the concept drifts exist or not. Online active learning has a generally accepted assumption that th...

show abstract

Online Passive-Aggressive Active learning

Cited by 67 publications

References 33 publications

Machine learning methods for optimal prediction of motor outcome in Parkinson’s disease

Machine learning methods for optimal prediction of motor outcome in Parkinson’s disease

Margin-Based Active Online Learning Techniques for Cooperative Spectrum Sharing in CR Networks

A Novel Sampling Strategy for Active Learning over Evolving Stream Data

Contact Info

Product

Resources

About