Online Active Learning Ensemble Framework for Drifted Data Streams

Gas sensors are the key components of an electronic nose (E-nose) in violated odour analysis. Gas-sensor drift is a kind of physical change on a sensor surface once an E-nose works. The perturbation of gas-sensor responses caused by drift would deteriorate the performance of the E-nose system over time. In this study, we intend to explore a suitable approach to deal with the drift effect in an online situation. Considering that the conventional drift calibration is difficult to implement online, we use active learning (AL) to provide reliable labels for online instances. Common AL learning methods tend to select and label instances with low confidence or massive information. Although this action clarifies the ambiguity near the classification boundary, it is inadequate under the influence of gas-sensor drift. We still need the samples away from the classification plane to represent drift variations comprehensively in the entire data space. Thus, a novel drift counteraction method named AL on adaptive confidence rule (AL-ACR) is proposed to deal with online drift data dynamically. By contrast with conventional AL methods selecting instances near the classification boundary of a certain category, AL-ACR collects instances distributed evenly in different categories. This action implements on an adjustable rule according to the outputs of classifiers. Compared with other reference methods, we adopt two drift databases of E-noses to evaluate the performance of the proposed method. The experimental results indicate that the AL-ACR reaches higher accuracy than references on two E-nose databases, respectively. Furthermore, the impact of the labelling number is discussed to show the trend of performance for the AL-type methods. Additionally, we define the labelling efficiency index (LEI) to assess the contribution of certain labelling numerically. According to the results of LEI, we believe AL-ACR can achieve the best effect with the lowest cost among the AL-type methods in this work.

show abstract

“…Considering the various variants of them, we use US proposed in Ref. [ 27 ], QBC and EQBC in Ref. [ 26 ].…”

Section: Methodsmentioning

confidence: 99%

Gas-Sensor Drift Counteraction with Adaptive Active Learning for an Electronic Nose

Liu

Chen

et al. 2018

Sensors

View full text Add to dashboard Cite

show abstract

“…Shan et al [22] also proposed an AL change detection strategy based on margin uncertainty, 'OALEnsemble', however in this approach the ensemble members are trained on different windows of the data set, with a stable classifier and a series of short window 'dynamic' classifiers that are continually replaced as new blocks of the data stream are processed, to balance the detection of both sudden and gradual concept drifts. Similar to [21], labeling is restricted to samples within the uncertainty margin, with the addition of a random labeling algorithm to randomly include samples outside of the margin where drift may also be occurring [22]. The stable classifier is incrementally trained with all new data, whilst dynamic classifiers are only trained on the most recent block and given a weight, providing importance to more recent data [22].…”

Section: Concept Drift Detection With Active Learningmentioning

confidence: 99%

SALAD: An Exploration of Split Active Learning based Unsupervised Network Data Stream Anomaly Detection using Autoencoders

Nixon¹,

Sedky²,

Hassan³

2021

Preprint

View full text Add to dashboard Cite

<div>Machine learning based intrusion detection systems monitor network data streams for cyber attacks. Challenges in this space include detection of unknown attacks, adaptation to changes in the data stream such as changes in underlying behaviour, the human cost of labeling data to retrain the machine learning model and the processing and memory constraints of a real-time data stream. Failure to manage the aforementioned factors could result in missed attacks, degraded detection performance, unnecessary expense or delayed detection times. This research evaluated autoencoders, a type of feed-forward neural network, as online anomaly detectors for network data streams. The autoencoder method was combined with an active learning strategy to further reduce labeling cost and speed up training and adaptation times, resulting in a proposed Split Active Learning Anomaly Detector (SALAD) method. The proposed method was evaluated with the NSL-KDD, KDD Cup 1999, and UNSW-NB15 data sets, using the scikit-multiflow framework. Results demonstrated that a novel Adaptive Anomaly Threshold method, combined with a split active learning strategy offered superior anomaly detection performance with a labeling budget of just 20%, significantly reducing the required human expertise to annotate the network data. Processing times of the autoencoder anomaly detector method were demonstrated to be significantly lower than traditional online learning methods, allowing for greatly improved responsiveness to attacks occurring in real time. Future research areas are applying unsupervised threshold methods, multi-label classification, sample annotation, and hybrid intrusion detection.</div>

show abstract

“…Compared with above single global modeling approaches, ensemble learning that employs multiple models to separately modeling the data subspaces has proven to be popular in online learning [6], [20]- [22]. In the area of selective ensemble learning, most of the researches focus on adaptive classification [23]- [25] and the regression problem is rarely discussed.…”

Section: Introductionmentioning

confidence: 99%

Growing and Pruning Selective Ensemble Regression for Nonlinear and Nonstationary Systems

et al. 2020

View full text Add to dashboard Cite

For a selective ensemble regression (SER) scheme to be effective in online modeling of fastarriving nonlinear and nonstationary data, it must not only be capable of maintaining a most up to date and diverse base model set but also be able to forget old knowledge no longer relevant. Based on these two important principles, in this paper, we propose a novel growing and pruning SER (GAP-SER) for timevarying nonlinear data. Specifically, during online operation, newly emerging process state is automatically identified and a local linear model is fitted to it. This adaptive growing strategy therefore maintains a most up to date and diverse local model set. The online prediction model is then constructed as a selective ensemble from the local linear model set based on a probability metric. Moreover, a pruning strategy is derived to remove 'unwanted' out of date local linear models in order to achieve low online computational complexity without sacrificing online modeling accuracy. A chaotic time series prediction and two real-world data sets are used to demonstrate the superior online modeling performance of the proposed GAP-SER over a range of benchmark schemes for nonlinear and nonstationary systems, in terms of online prediction accuracy and computational complexity. INDEX TERMS Nonlinear and nonstationary data, local linear model, growing model, pruning model, selective ensemble.

show abstract

Online Active Learning Ensemble Framework for Drifted Data Streams

Cited by 97 publications

References 34 publications

Gas-Sensor Drift Counteraction with Adaptive Active Learning for an Electronic Nose

Gas-Sensor Drift Counteraction with Adaptive Active Learning for an Electronic Nose

SALAD: An Exploration of Split Active Learning based Unsupervised Network Data Stream Anomaly Detection using Autoencoders

Growing and Pruning Selective Ensemble Regression for Nonlinear and Nonstationary Systems

Contact Info

Product

Resources

About