The main task in causal inference is predicting the outcome of an intervention. For example, a treatment assigned by a doctor that will change the patient’s heart condition is an intervention. Predicting the change in the patient’s condition is a causal inference task. In general, an intervention is an action taken by an external agent that changes the original values, or the probability distributions, of some of the variables in the system. Besides predicting outcomes of actions, causal inference is also concerned with explanation: identifying which were the causes of a particular event that happened in the past.
Motivation and Background
Many problems in machine learning are prediction problems. Given a feature vector X, the task is to provide an estimate of some output vector Y, or its conditional probability distribution P(Y | X). This typically assumes that the distribution of Y given Xduring learning is the same distribution at prediction time. There are many...
- Cooper, G., & Yoo, C. (1999). Causal discovery from a mixture of experimental and observational data. In Uncertainty in Artificial Intelligence (UAI).Google Scholar
- Eaton, D., & Murphy, K. (2007). Exact Bayesian structure learning from uncertain interventions. In Artificial Intelligence and Statistics (AISTATS).Google Scholar
- Eberhardt, F., Glymour, C., & Scheines, R. (2005). On the number of experiments sufficient and in the worst case necessary to identify all causal relations among N variables. In Uncertainty in Artificial Intelligence (UAI) (pp. 178–184).Google Scholar
- Halpern, J., & Pearl, J. (2005). Causes and explanations: A structural-model approach. Part II: Explanations. British Journal for the Philosophy of Science, 56, 889–911.Google Scholar
- Meek, C. (1997). Graphical models: Selecting causal and statistical models. PhD thesis, Carnegie Mellon University.Google Scholar
- Richardson, T. (1996). A discovery algorithm for directed cyclic graphs. In Proceedings of 12th conference on Uncertainty in Artificial Intelligence.Google Scholar
- Sachs, K., Prez, O., Pe’er, D., Lauffenburger, D., & Nolan, G. (2005). Causal protein-signaling networks derived from multiparameter single-cell data. Science, 308, 523–529.Google Scholar
- Scheines, R. (1997). An introduction to causal inference. In V. McKim & S. Turner (Eds.), Causality in Crisis? (pp. 185–200).Google Scholar
- Spirtes, P., Glymour, C., & Scheines, R. (2000). Causation, prediction and search. Cambridge, MA: Cambridge University Press.Google Scholar
- Zhang, J., & Spirtes, P. (2003). Strong faithfulness and uniform consistency in causal inference. In Uncertainty in Artificial Intelligence.Google Scholar