| title | layout |
|---|---|
OCCAM - What Is Reconstructibility Analysis? |
default |
OCCAM is software developed at Portland State University under the supervision of Dr. Martin Zwick. It implements Reconstructibility Analysis (RA), a method developed in the systems community, and which uses information theory to:
- search for possible models of a set of data
- choose between those models in a theoretically well-justified way
- fit a chosen model to the data, and then
- use the model for understanding the data, analyzing relationships between variables and groups of variables, making predictions, and in general gaining deep insight into the structure of the data and the ways it can be put to use.
The theoretical background of RA is not accessible to casual users, but should look familiar to those who have experience in machine learning, data mining, statistical analysis, and related areas. Fortunately, OCCAM makes it possible to take advantage of the machinery of RA without detailed knowledge of the underlying theory.
For in-depth introductions to RA, see:
- An Overview of Reconstructibility Analysis{:target="_blank"}
- Wholes and Parts In General Systems Methodology{:target="_blank"}
For more on applications and past and present research using OCCAM, see the RA and OCCAM research page at Portland State University.