Algorithms for Estimating Information Distance

Title: Algorithms for Estimating Information Distance

Research Question: How can we estimate the information distance between two objects using data compression algorithms?

Methodology: The researchers proposed two methods to estimate the information distance between two objects. The first method is based on the normalized information distance (E2(x,y)), which was proposed by Li et al. The second method is based on data compression algorithms, which are used to approximate the Kolmogorov complexities.

Results: The researchers showed that the data compression algorithms can be used to estimate the entropy rates and conditional entropy rates, which are used to calculate the information distance. They also found that the second method provides a good approximation of the normalized information distance for sufficiently large data sets.

Implications: The results of this research have implications for the field of bioinformatics and computational linguistics, where the information distance between objects is often used to measure similarity or dissimilarity. The proposed methods can be used to develop more accurate and efficient algorithms for estimating the information distance between objects.

Link to Article: https://arxiv.org/abs/0404039v1 Authors: arXiv ID: 0404039v1