Criticality of mostly informative samples: A Bayesian model selection approach

Abstract

We discuss a Bayesian model selection approach to high dimensional data in the deep under sampling regime. The data is based on a representation of the possible discrete states $s$, as defined by the observer, and it consists of $M$ observations of the state. This approach shows that, for a given sample size $M$, not all states observed in the sample can be distinguished. Rather, only a partition of the sampled states $s$ can be resolved. Such partition defines an {\em emergent} classification $q_s$ of the states that becomes finer and finer as the sample size increases, through a process of {\em symmetry breaking} between states. This allows us to distinguish between the $resolution$ of a given representation of the observer defined states $s$, which is given by the entropy of $s$, and its $relevance$ which is defined by the entropy of the partition $q_s$. Relevance has a non-monotonic dependence on resolution, for a given sample size. In addition, we characterise most relevant samples and we show that they exhibit power law frequency distributions, generally taken as signatures of "criticality". This suggests that "criticality" reflects the relevance of a given representation of the states of a complex system, and does not necessarily require a specific mechanism of self-organisation to a critical point.

31 pages, 7 figures

References in corpus (3)

• Power-law distributions in empirical data
• Scale-free brain functional networks
• Identifying relevant positions in proteins by Critical Variable Selection

Cited by in corpus (18)

• Colloquium: Criticality and dynamical scaling in living systems
• Statistical Criticality arises in Most Informative Representations
• Does the brain behave like a (complex) network? I. Dynamics
• Quantifying Relevance in Learning and Inference
• Identifying relevant positions in proteins by Critical Variable Selection
• Minimum Description Length codes are critical
• Ensemble Inhibition and Excitation in the Human Cortex: an Ising Model Analysis with Uncertainties
• Making sense of complex systems through resolution, relevance, and mapping entropy
• Dynamical signatures of structural connectivity damage to a model of the brain posed at criticality
• On Maximum Entropy and Inference
• Improving randomness characterization through Bayesian model selection
• Scale-invariant representation of machine learning
• A New Spin on Color Quantization
• Resolution and Relevance Trade-offs in Deep Learning
• Maximal Relevance and Optimal Learning Machines
• Multiscale Relevance of Natural Images
• Filtering Statistics on Networks
• Complex distributions emerging in filtering and compression