Archive - Central European Conference on Information and Intelligent Systems, CECIIS - 2011

Font Size: 
Causal models
Marcel Maretić

Last modified: 2011-10-03

Abstract


We present a framework for causal modeling. Main objective is to define models augmented with graphs and to present relevant results and important concepts in the theory of structural causal models. Causal models assume autonomy of mechanisms involved. This feature allows us to predict their behaviour even when some of these mechanisms break or change. We focus our attention on Markovian and semi-Markovian models. Causal effect is defined and we show how and when it can be calculated.

Full Text: PDF