New Information-Theory-Based Methods in the Analysis of Childhood Development Data
The complexity of infant growth and development processes and resultant data reflect the deep complexity of biological systems.
The Problem: Can we detect complex dependencies of biological variables?
Given data sets with many instances of many variables:
- Can we find a measure for a subset of variables that is significantly nonzero if and only if the variables are interdependent?
- Can we separate detection of the existence of dependence from any models of the nature of dependence?