A benchmark and comparison of active learning for logistic regression

Yang, Yazhou; Loog, Marco

doi:10.1016/j.patcog.2018.06.004

Cited by 126 publications

(105 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We focus on logistic regression because it has been widely used in applied sciences [11]. We focus on the binary classification case in which there are two classes.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Learning with Vertically-Partitioned Data, Binary Feedback, and Random Parameter Update

Nguyen

Sigg

2019

IEEE INFOCOM 2019 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)

View full text Add to dashboard Cite

Machine learning models can deal with data samples scattered among distributed agents, each of which holds a nonoverlapping set of sample features. In this paper, we propose a training algorithm that does not require communication between these agents. A coordinator can access ground-truth labels and produces binary feedback to guide the optimization process towards optimal model parameters. We mimic the gradient descent technique with information observed locally at each agent. We experimented with the logistic regression model on multiple benchmark datasets and achieves promising results in terms of convergence rate and communication load.

show abstract

“…We focus on logistic regression because it has been widely used in applied sciences [11]. We focus on the binary classification case in which there are two classes.…”

Section: Methodsmentioning

confidence: 99%

“…In our comparison on the KDD99 dataset [15], J = 1, C = 2, and we varied B. It has been used to as a standard benchmark dataset for classification algorithms, for example in [5] and [11]. Note that we aim to classify normal and abnormal cases (attacking), hence C = 2.…”

Section: B On Amount Of Transmitted Datamentioning

confidence: 99%

Learning with Vertically-Partitioned Data, Binary Feedback, and Random Parameter Update

Nguyen

Sigg

2019

IEEE INFOCOM 2019 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS)

View full text Add to dashboard Cite

show abstract

“…A critical aspect for an active learner is represented by the strategy used to query the next sample to be labeled. Four main query frameworks exist, which rely mostly on heuristics: informativeness [58,13,17,4], representativeness [46,48], hybrid [22,57], and performance-based [47,16,12,56]. Among all these, informativeness-based approaches are the most successful ones.…”

Section: Related Workmentioning

confidence: 99%

Temporal Coherence for Active Learning in Videos

Bengar

González-García²,

Villalonga

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

View full text Add to dashboard Cite

Autonomous driving systems require huge amounts of data to train. Manual annotation of this data is timeconsuming and prohibitively expensive since it involves human resources. Therefore, active learning emerged as an alternative to ease this effort and to make data annotation more manageable. In this paper, we introduce a novel active learning approach for object detection in videos by exploiting temporal coherence. Our active learning criterion is based on the estimated number of errors in terms of false positives and false negatives. The detections obtained by the object detector are used to define the nodes of a graph and tracked forward and backward to temporally link the nodes. Minimizing an energy function defined on this graphical model provides estimates of both false positives and false negatives. Additionally, we introduce a synthetic video dataset, called SYNTHIA-AL, specially designed to evaluate active learning for video object detection in road scenes. Finally, we show that our approach outperforms active learning baselines tested on two datasets.

show abstract

“…The problem formulation has direct relations to sequential analysis [121] and optimal experimental design [40]. Overviews of current techniques can be found in [19], [111], and [126]. One of the major issues in active learning is that the systematic collection of labeled training data typically leads to a systematic bias as well.…”

Section: Active Learningmentioning

confidence: 99%

Supervised Classification: Quite a Brief Overview

Loog

2018

Machine Learning Techniques for Space Weather

Self Cite

View full text Add to dashboard Cite

The original problem of supervised classification considers the task of automatically assigning objects to their respective classes on the basis of numerical measurements derived from these objects. Classifiers are the tools that implement the actual functional mapping from these measurements-also called features or inputs-to the so-called class label-or output. The fields of pattern recognition and machine learning study ways of constructing such classifiers. The main idea behind supervised methods is that of learning from examples: given a number of example input-output relations, to what extent can the general mapping be learned that takes any new and unseen feature vector to its correct class? This chapter provides a basic introduction to the underlying ideas of how to come to a supervised classification problem. In addition, it provides an overview of some specific classification techniques, delves into the issues of object representation and classifier evaluation, and (very) briefly covers some variations on the basic supervised classification task that may also be of interest to the practitioner.

show abstract

A benchmark and comparison of active learning for logistic regression

Cited by 126 publications

References 53 publications

Learning with Vertically-Partitioned Data, Binary Feedback, and Random Parameter Update

Learning with Vertically-Partitioned Data, Binary Feedback, and Random Parameter Update

Temporal Coherence for Active Learning in Videos

Supervised Classification: Quite a Brief Overview

Contact Info

Product

Resources

About