Linkage and Autocorrelation Cause Feature Selection Bias in Relational Learning

By David D. Jensen, Jennifer Neville. 2002

202 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Fitted Q-Learning for Relational Domains (2020, background): “…a ﬂat feature vector from a symbolic structure to train a deep model can lead to loss of information including auto-correlation and linkage (Jensen and Neville 2002 Results: To answer Q1 , it can be observed from the re-sults reported in Figure 5 that the proposed GBQL and RBFQ methods…”
A Review of Relational Machine Learning for Knowledge Graphs (2015, background): “For multirelational data (graphs with more than one kind of link), homophily has also been referred to as autocorrelation [57].”
A Simple Relational Classifier (2003, background): “Recent work has begun to investigate foundational issues within relational learning, such as the dimensions across which learners can be compared [11, 14, 25] as well as issues of link dependencies [13].”
Relational Dependency Networks (2007, methodology): “…2001; Neville and Jensen, 2003) • Functions of colocated proteins in a cell (Neville and Jensen, 2002) • Box-office receipts of movies made by the same studio (Jensen and Neville, 2002) • Industry categorization of corporations that co-occur in news stories (Bernstein et al., 2003) •…”
Collective Classification in Network Data (2008, background): “Even though collective classification has gained attention only in the past five to seven years, initiated by the work of Jennifer Neville and David Jensen [44, 24, 25, 46] and the work of Ben Taskar et al.”
Statistical Relational Learning for Link Prediction (2003, background): “The assumption that the examples are independent is violated in the presence of relational structure ; this can be addressed explicitly [Jensen and Neville, 2002; Hoff, 2003], or implicitly, as we do here, by generating more complex features which capture relational dependencies .”
Feature Selection for Classification: A Review (2014, background): “), which is among the most enduring and deeply buried assumptions of traditional machine learning methods [31, 63].”
A bias/variance decomposition for models using collective inference (2008, background): “Relational autocorrelation is a statistical dependency between the values of the same variable (e.g., the class label) on related entities and is a nearly ubiquitous characteristic of relational datasets (Jensen and Neville 2002; Taskar et al. 2002; Bernstein et al. 2003; Hill et al. 2006).”
Modeling relationship strength in online social networks (2010, background): “In relational data, autocorrelation is the statistical dependency of the same attribute on related instances [10].”
Classification in Networked Data: A Toolkit and a Univariate Case Study (2007, unknown): “The data are patently not independent and identically distributed, which introduces bias to learning and inference procedures (Jensen and Neville, 2002b).”
Why collective inference improves relational classification (2004, background): “Collective inference can exploit relational autocorrelation, a widely observed characteristic of relational data in which the value of a variable for one instance is highly correlated with the value of the same variable on another instance [7].”
A Sound and Complete Algorithm for Learning Causal Models from Relational Data (2013, background): “If RBO is akin to detecting relational autocorrelation (Jensen and Neville, 2002) and checking whether a distinct variable is a member of the set that eliminates the autocorrelation.”
Feature Selection for Clustering: A Review (2018, background): “), which is among the most enduring and deeply buried assumptions of traditional machine learning methods [29, 66].”
Leveraging relational autocorrelation with latent group models (2005, background): “Autocorrelation is a statistical dependency between the values of the same variable on related entities and is a nearly ubiquitous characteristic of relational data sets.”