Feature Space and Label Space Selection Based on Error-Correcting Output Codes for Partial Label Learning

Partial label learning aims to learn from training examples each associated with a set of candidate labels, among which only one label is valid for the training example. The common strategy to induce predictive model is trying to disambiguate the candidate label set, such as disambiguation by identifying the ground-truth label iteratively or disambiguation by treating each candidate label equally. Nonetheless, these strategies ignore considering the generalized label distribution corresponding to each instance since the generalized label distribution is not explicitly available in the training set. In this paper, a new partial label learning strategy named PL-LE is proposed to learn from partial label examples via label enhancement. Specifically, the generalized label distributions are recovered by leveraging the topological information of the feature space. After that, a multi-class predictive model is learned by fitting a regularized multi-output regressor with the generalized label distributions. Extensive experiments show that PL-LE performs favorably against state-ofthe-art partial label learning approaches.

Download Full-text

Disambiguation Enabled Linear Discriminant Analysis for Partial Label Dimensionality Reduction

ACM Transactions on Knowledge Discovery from Data ◽

10.1145/3494565 ◽

2022 ◽

Vol 16 (4) ◽

pp. 1-18

Author(s):

Min-Ling Zhang ◽

Jing-Han Wu ◽

Wei-Xuan Bao

Keyword(s):

Discriminant Analysis ◽

Dimensionality Reduction ◽

Linear Discriminant Analysis ◽

Feature Space ◽

Ground Truth ◽

Learning System ◽

Generalization Performance ◽

Linear Discriminant ◽

Training Examples ◽

Partial Label Learning

As an emerging weakly supervised learning framework, partial label learning considers inaccurate supervision where each training example is associated with multiple candidate labels among which only one is valid. In this article, a first attempt toward employing dimensionality reduction to help improve the generalization performance of partial label learning system is investigated. Specifically, the popular linear discriminant analysis (LDA) techniques are endowed with the ability of dealing with partial label training examples. To tackle the challenge of unknown ground-truth labeling information, a novel learning approach named Delin is proposed which alternates between LDA dimensionality reduction and candidate label disambiguation based on estimated labeling confidences over candidate labels. On one hand, the (kernelized) projection matrix of LDA is optimized by utilizing disambiguation-guided labeling confidences. On the other hand, the labeling confidences are disambiguated by resorting to k NN aggregation in the LDA-induced feature space. Extensive experiments over a broad range of partial label datasets clearly validate the effectiveness of Delin in improving the generalization performance of well-established partial label learning algorithms.

Download Full-text

Partial label learning based on label distributions and error-correcting output codes

Soft Computing ◽

10.1007/s00500-020-05203-0 ◽

2020 ◽

Author(s):

Guangyi Lin ◽

Kunhong Liu ◽

Beizhan Wang ◽

Xiaoyan Zhang

Keyword(s):

Partial Label Learning ◽

Error Correcting Output Codes

Download Full-text

Discriminative and Correlative Partial Multi-Label Learning

Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence ◽

10.24963/ijcai.2019/512 ◽

2019 ◽

Cited By ~ 4

Author(s):

Haobo Wang ◽

Weiwei Liu ◽

Yang Zhao ◽

Chen Zhang ◽

Tianlei Hu ◽

...

Keyword(s):

Real World ◽

State Of The Art ◽

Feature Space ◽

Gradient Boosting ◽

Training Procedure ◽

Second Stage ◽

Label Correlations ◽

Real World Datasets ◽

Partial Label Learning ◽

Confidence Value

In partial label learning (PML), each instance is associated with a candidate label set that contains multiple relevant labels and other false positive labels. The most challenging issue for the PML is that the training procedure is prone to be affected by the labeling noise. We observe that state-of-the-art PML methods are either powerless to disambiguate the correct labels from the candidate labels or incapable of extracting the label correlations sufficiently. To fill this gap, a two-stage DiscRiminative and correlAtive partial Multi-label leArning (DRAMA) algorithm is presented in this work. In the first stage, a confidence value is learned for each label by utilizing the feature manifold, which indicates how likely a label is correct. In the second stage, a gradient boosting model is induced to fit the label confidences. Specifically, to explore the label correlations, we augment the feature space by the previously elicited labels on each boosting round. Extensive experiments on various real-world datasets clearly validate the superiority of our proposed method.

Download Full-text

The Utility of an Intermediate Representation of Feature Space: Lessons From Fingerprint Examiners

PsycEXTRA Dataset ◽

10.1037/e502412013-026 ◽

2012 ◽

Author(s):

Tom Busey ◽

Chen Yu ◽

Francisco Parada ◽

Brandi Emerick ◽

John Vanderkolk

Keyword(s):

Feature Space ◽

Intermediate Representation

Download Full-text

Context-Driven Proactive Decision Support for Hybrid Teams

AI Magazine ◽

10.1609/aimag.v40i3.4810 ◽

2019 ◽

Vol 40 (3) ◽

pp. 41-57

Author(s):

Manisha Mishra ◽

Pujitha Mannaru ◽

David Sidoti ◽

Adam Bienkowski ◽

Lingyi Zhang ◽

...

Keyword(s):

Decision Support ◽

Situational Awareness ◽

Feature Space ◽

Human Interaction ◽

Superior Performance ◽

Uncertain Environments ◽

Cognitive States ◽

Contextual Elements ◽

Human Operators ◽

Smart Machine

A synergy between AI and the Internet of Things (IoT) will significantly improve sense-making, situational awareness, proactivity, and collaboration. However, the key challenge is to identify the underlying context within which humans interact with smart machines. Knowledge of the context facilitates proactive allocation among members of a human–smart machine (agent) collective that balances autonomy with human interaction, without displacing humans from their supervisory role of ensuring that the system goals are achievable. In this article, we address four research questions as a means of advancing toward proactive autonomy: how to represent the interdependencies among the key elements of a hybrid team; how to rapidly identify and characterize critical contextual elements that require adaptation over time; how to allocate system tasks among machines and agents for superior performance; and how to enhance the performance of machine counterparts to provide intelligent and proactive courses of action while considering the cognitive states of human operators. The answers to these four questions help us to illustrate the integration of AI and IoT applied to the maritime domain, where we define context as an evolving multidimensional feature space for heterogeneous search, routing, and resource allocation in uncertain environments via proactive decision support systems.

Download Full-text