Generalized Expectation Criteria for Semi-Supervised Learning with Weakly Labeled Data

By Gideon S. Mann, Andrew McCallum. 2010

221 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

Snorkel: Rapid Training Data Creation with Weak Supervision (2017, background): “Other forms include crowdsourced labels [37, 50], rules and heuristics for labeling data [39, 52], and others [29, 30, 30, 46, 51].”
Joint Concept Learning and Semantic Parsing from Natural Language Explanations (2017, methodology): “Notable approaches such as Generalized Expectation (Mann and McCallum, 2010) and Posterior Regularization (Ganchev et al., 2010) have explored integration of manually provided ‘side-information’ (feature and label constraints) to guide machine learning models.”
Concrete Problems in AI Safety (2016, background): “e noisy hints about the correct evaluations There has been some work in this direction within the area of semi-supervised or weakly supervised learning. For instance, generalized expectation criteria [94, 45] ask the user to provide population-level statistics (e.g. telling the system that on averag”
Optimization with Non-Differentiable Constraints with Applications to Fairness, Recall, Churn, and Other Goals (2018, methodology): “Mann and McCallum (2007) and follow-on work (Bellare et al., 2009; Mann and McCallum, 2010) optimized probabilistic models with constraints in order to incorporate side information about the prior priors on class labels, which in the context of a binary classifier is a special case of rate constrain”
Active Learning (2012, background): “, 2011), and even more complex graphical models (Ganchev et al., 2010; Mann and McCallum, 2010).”
From Theories to Queries (2011, methodology): “Mann and McCallum (2010) describe a semi-supervised method of combining such constraints with unlabeled data in exponential models, and Melville et al. (2009) combine this domain knowledge with labeled examples for näıve Bayes by pooling multinomials.”
Learning Classifiers from Declarative Language (2018, background): “Most prominent among these are the Generalized Expectation [8] and Posterior Regularization [3] frameworks that integrate manually provided ‘sideinformation’ (feature and label constraints) to guide machine learning models.”
Tractable Semi-supervised Learning of Complex Structured Prediction Models (2013, methodology): “Several semi-supervised learning algorithms [2,23,36,40,1,34,7] have been proposed in the literature.”
Training Complex Models with Multi-Task Weak Supervision (2018, methodology): “2018), feature annotations (Mann and McCallum 2010; Zaidan and Eisner 2008), and noisy crowd labels (Karger, Oh, and Shah 2011; Dawid and Skene 1979).”
Code Mixing: A Challenge for Language Identification in the Language of Social Media (2014, methodology): “and sequence labelling with Conditional Random Fields trained with Generalized Expectation criteria (Mann and McCallum, 2008; Mann and McCallum, 2010), which achieved the highest scores.”
Infinite Latent SVM for Classification and Multi-task Learning (2011, methodology): “For the first step, we can use sampling or variational methods [9] to do approximate inference; and under certain conditions, such as using the constraints based on posterior expectation [21], the second step can be efficiently done using high-performance convex optimization techniques, as we shall ”
DataComp: In search of the next generation of multimodal datasets (2023, background): “[89] Gideon S Mann and Andrew McCallum.”
A Survey on Programmatic Weak Supervision (2022, background): “There are also other options, including crowd-sourced labels [ Yuen et al. , 2011 ] , heuristic rules [ Awasthi et al. , 2020 ] , feature annotation [ Mann and McCallum, 2010 ] , and others.”
Zero-shot Learning of Classifiers from Natural Language Quantification (2018, methodology): “Prominent among these are the Constraint-driven learning (Chang et al., 2007a), Generalized Expectation (Mann and McCallum, 2010) and Posterior Regularization (Ganchev et al., 2010) and Bayesian Measurements (Liang et al., 2009) frameworks.”
Weakly supervised learning via statistical sufficiency (2016, methodology): “…of a highly rated movie review; a sample of those methods are constraint driven learning Chang et al. [2007], generalized expectation constraints Mann and McCallum [2008], posterior constraints Ganchev et al. [2010], and label regularization [Mann and McCallum, 2010; Ardehaly and Culotta, 2015].”
Large Scale Text Classification using Semisupervised Multinomial Naive Bayes (2011, methodology): “(Mann & McCallum, 2010) proposed the Generalized Expectation method and observed that the classical EM+MNB outperforms it in text classification datasets.”