Advances in Minimum Description

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

Deep learning in neural networks: An overview (2014, methodology): “Given some programming language, the principle of Minimum Description Length (MDL) can be used to measure the complexity of a solution candidate by the length of the shortest program that computes it (e.g., Solomonoff, 1964; Kolmogorov, 1965b; Chaitin, 1966; Wallace and Boulton, 1968; Levin, 1973a; ”
On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models (2015, methodology): “Instead, to avoid the erroneous “discovery” of “regular patterns” in irregular noise, we use AIT’s sound way of dealing with overfitting [263, 130, 289, 207, 147, 84], and measure M ’s compression performance by the number of bits required to specify M , plus the bits needed to encode the observed d”
Community detection in graphs (2009, methodology): “Chakrabarti (Chakrabarti, 2004) has applied the MDL principle to put the adjacency matrix of a graph into the (approximately) block diagonal form representing the best tradeoff between having a limited number of blocks, for a good compression of the graph topology, and having very homogeneous blocks”
‘All models are wrong...’: an introduction to model uncertainty (2012, background): “This is a non-trivial task as the integration in Equation (3) for moderately complex models can be across a large sample space M.”
Multilevel Compression of Random Walks on Networks Reveals Hierarchical Organization in Large Integrated Systems (2010, methodology): “n networks. This approach uses a random walk as a proxy for the real ow (14,15), and exploits the duality between compressing a message and nding patterns in the structure that generates that message (16,17). To nd the regularities that induce the dynamics on networks, the map equation measures, for”
Trajectory Outlier Detection: A Partition-and-Detect Framework (2008, background): “The two components are informally stated as follows [16]: “L(H) is the length, in bits, of the description of the hypothesis; and L(D|H) is the length, in bits, of the description of the data when encoded with the help of the hypothesis.”
Community detection in networks: A user guide (2016, background): “A possibility is model selection, for instance by choosing the model that best compresses the data [139,140].”
Genetic process mining: an experimental evaluation (2007, methodology): “In contrast to the previous papers, the work (Maruster et al. 2002; Weijters and van der Aalst 2003) is characterized by the focus on workflow processes with concurrent behavior (rather than adding ad-hoc mechanisms to capture parallelism). In Weijters and van der Aalst (2003), a heuristic approach ”
Learning action models from plan examples using weighted MAX-SAT (2007, methodology): “…states are discrete-valued, then the problem is one of predicting a minimal set of actions and the associated models, such that the number of actions, sizes of precondition and effect sets, as well as the model’s error rate, are minimized according to the minimal description length principle [20].”
Model Comparison and the Principle of Parsimony (2015, methodology): “In psychology, the MDL principle has been applied and promoted primarily by Grünwald (2000), Grünwald, Myung, and Pitt (2005), and Grünwald (2007), as well as Myung, Navarro, and Pitt (2006), Pitt and Myung (2002), and Pitt, Myung, and Zhang (2002).”
A Tutorial on Fisher Information (2017, background): “…a measure of model15 complexity and its role in model selection within the minimum description length framework16 (MDL; de Rooij and Grünwald, 2011; Grünwald et al., 2005; Grünwald, 2007; Myung et al.,17 2000b; Myung et al., 2006; Pitt et al., 2002).18 The primary aim of a model selection…”
Community detection algorithms: a comparative analysis: invited presentation, extended abstract (2009, result): “1 we show the results of our analysis.”
An information-theoretic framework for resolving community structure in complex networks (2006, methodology): “This is an optimal coding problem and can be resolved by the Minimum Description Length (MDL) principle (12, 21, 22).”
Maps of random walks on complex networks reveal community structure (2007, methodology): “). Taking this approach, we develop an efﬁcient code to describe a random walk on a network (14). We thereby show that ﬁnding community structure in networks is equivalent to solving a coding problem (6, 15, 16). We exemplify this by making a map of science, based on how information ﬂows among scien”
Trajectory clustering: a partition-and-group framework (2007, methodology): “For the .rst phase, we present a formal trajectory partitioning algorithm using the minimum description length (MDL) principle.”
Learning Phonology With Substantive Bias: An Experimental and Computational Study of Velar Palatalization (2006, background): “L Λ = [− N j=1 log P Λ (y (j) |x (j) )] − [ K k=1 (λ k − µ k ) 2 ) 2σ 2 k ] (6) This equation defines the likelihood of the weights (L Λ ) as a function of two bracketed terms, which have interpretations that are familiar from the theory of induction (e.g., Grünwald et al., 2005; Smolensky, 1996).”
Predictive Feedback Can Account for Biphasic Responses in the Lateral Geniculate Nucleus (2009, background): “MDL chooses as the best model the one that enables the shortest code length for both prediction error and model parameters in bits (in base e) [77,78].”