Aggregation and Population Growth : The Relational Logistic Regression and Markov Logic Cases

This paper considers how relational probabilistic models adapt to population size. First we show that what are arbitrary choices for nonrelational domains become a commitment to how a relational model adapts to population change. We show how this manifests in a directed model where the conditional probabilities are represented using the logistic function, and show why it needs to be extended to a relational logistic function. Second we prove that directed aggregation models cannot be represented by Markov Logic without clauses that involve multiple individuals. Third we show how these models change as a function of population size.

[1]  Kristian Kersting,et al.  Exploiting Causal Independence in Markov Logic Networks: Combining Undirected and Directed Models , 2010, StarAI@AAAI.

[2]  Michael Beetz,et al.  Adaptive Markov Logic Networks: Learning Statistical Relational Models with Dynamic Parameters , 2010, ECAI.

[3]  Luc De Raedt,et al.  Probabilistic Inductive Logic Programming , 2004, Probabilistic Inductive Logic Programming.

[4]  Matthew Richardson,et al.  Markov Logic , 2008, Probabilistic Inductive Logic Programming.

[5]  Michael Beetz,et al.  Extending Markov Logic to Model Probability Distributions in Relational Domains , 2007, KI.

[6]  Matthew Richardson,et al.  Markov logic networks , 2006, Machine Learning.

[7]  Foster J. Provost,et al.  Distribution-based aggregation for relational learning with identifier attributes , 2006, Machine Learning.

[8]  Jennifer Neville,et al.  Using relational knowledge discovery to prevent securities fraud , 2005, KDD '05.

[9]  Dan Roth,et al.  Lifted First-Order Probabilistic Inference , 2005, IJCAI.

[10]  David Poole,et al.  First-order probabilistic inference , 2003, IJCAI.

[11]  Lise Getoor,et al.  Learning Probabilistic Relational Models , 1999, IJCAI.

[12]  David Poole,et al.  A Dynamic Approach to Probabilistic Inference using Bayesian Networks , 1990, UAI 1990.

[13]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems - networks of plausible inference , 1991, Morgan Kaufmann series in representation and reasoning.