Learning Markov Logic Networks via Functional Gradient Boosting

By Tushar Khot, Sriraam Natarajan, Kristian Kersting, Jude W. Shavlik. 2011

108 citations recorded by the source corpus.

Recent years have seen a surge of interest in Statistical Relational Learning (SRL) models that combine logic with probabilities. One prominent example is Markov Logic Networks (MLNs). While MLNs are indeed highly expressive, this expressiveness comes at a cost. Learning MLNs is a hard problem and therefore has attracted much interest in the SRL community. Current methods for learning MLNs follow a two-step approach: first, perform a search through the space of possible clauses and then learn appropriate weights for these clauses. We propose to take a different approach, namely to learn both the weights and the structure of the MLN simultaneously. Our approach is based on functional gradient boosting where the problem of learning MLNs is turned into a series of relational functional approximation problems. We use two kinds of representations for the gradients: clause-based and tree-based. Our experimental evaluation on several benchmark data sets demonstrates that our new approach can learn MLNs as good or better than those found with state-of-the-art methods, but often in a fraction of the time.

Subfields: Artificial Intelligence, Information Systems, Management Science and Operations Research

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 Information Systems

In Management Science and Operations Research

Citations in the corpus

Fitted Q-Learning for Relational Domains (2020, methodology): “To represent these lifted representations, as mentioned earlier, we use RFGB (Natarajan et al. 2014; Natarajan et al. 2012; Natarajan et al. 2011) and a single tree, TILDE (Blockeel and De Raedt 1998), for approximating the Q-value functions.”
Transfer Learning via Relational Type Matching (2015, background): “While there is significant research in learning these parameterized rules, especially for MLNs ( [12] ), these methods assume the presence of large amounts of training data.”
Human-Guided Learning for Probabilistic Logic Models (2018, methodology): “Also, it answers Q4 that advice can be used to learn MLNs.”
Structure learning for relational logistic regression: an ensemble approach (2018, methodology): “The key difference to the existing gradient boosting methods for RDNs [32] and MLNs [17] is that the RLR learning algorithm needs to learn a weight vector per clause instead of a single weight.”
Statistical Relational Artificial Intelligence: Logic, Probability, and Computation (2016, methodology): “Recently, a boosted approach to learning MLNs has been developed [Khot et al., 2011] and we cover this later in the chapter.”