Challenges and Opportunities with Causal Discovery Algorithms: Application to Alzheimer’s Pathophysiology

It is commonplace to encounter nonstationary or heterogeneous data, of which the underlying generating process changes over time or across data sets (the data sets may have different experimental conditions or data collection conditions). Such a distribution shift feature presents both challenges and opportunities for causal discovery. In this paper we develop a principled framework for causal discovery from such data, called Constraint-based causal Discovery from Nonstationary/heterogeneous Data (CD-NOD), which addresses two important questions. First, we propose an enhanced constraint-based procedure to detect variables whose local mechanisms change and recover the skeleton of the causal structure over observed variables. Second, we present a way to determine causal orientations by making use of independence changes in the data distribution implied by the underlying causal model, benefiting from information carried by changing distributions. Experimental results on various synthetic and real-world data sets are presented to demonstrate the efficacy of our methods.

Download Full-text

Comment on "Estimating causal networks in biosphere–atmosphere interaction with the PCMCI approach"

10.5194/bg-2021-231 ◽

2021 ◽

Author(s):

Jarmo Mäkelä ◽

Laila Melkas ◽

Ivan Mammarella ◽

Tuomo Nieminen ◽

Suyog Chandramouli ◽

...

Keyword(s):

Domain Knowledge ◽

Prior Information ◽

Causal Model ◽

Concept Drift ◽

Causal Structure ◽

Causal Discovery ◽

Causal Networks ◽

Atmosphere Interaction ◽

Model Structures ◽

Discovery Algorithms

Abstract. This is a comment on "Estimating causal networks in biosphere–atmosphere interaction with the PCMCI approach" by Krich et al., Biogeosciences, 17, 1033–1061, 2020, which gives a good introduction to causal discovery, but conﬁnes the scope by investigating the outcome of a single algorithm. In this comment, we argue that the outputs of causal discovery algorithms should not usually be considered as end results but starting points and hypothesis for further study. We illustrate how not only different algorithms, but also different initial states and prior information of possible causal model structures, affect the outcome. We demonstrate how to incorporate expert domain knowledge with causal structure discovery and how to detect and take into account overﬁtting and concept drift.

Download Full-text

Comparative benchmarking of causal discovery algorithms

Proceedings of the ACM India Joint International Conference on Data Science and Management of Data - CoDS-COMAD '18 ◽

10.1145/3152494.3152499 ◽

2018 ◽

Cited By ~ 1

Author(s):

Karamjit Singh ◽

Garima Gupta ◽

Vartika Tewari ◽

Gautam Shroff

Keyword(s):

Causal Discovery ◽

Discovery Algorithms

Download Full-text

Using Causal Discovery Algorithms to Learn About Our Planet’s Climate

Machine Learning and Data Mining Approaches to Climate Science ◽

10.1007/978-3-319-17220-0_11 ◽

2015 ◽

pp. 113-126 ◽

Cited By ~ 2

Author(s):

Imme Ebert-Uphoff ◽

Yi Deng

Keyword(s):

Causal Discovery ◽

Discovery Algorithms

Download Full-text

Identification of linkages between urban heat Island magnitude and urban rainfall modification by use of causal discovery algorithms

Urban Climate ◽

10.1016/j.uclim.2020.100659 ◽

2020 ◽

Vol 33 ◽

pp. 100659

Author(s):

Jie Liu ◽

Dev Niyogi

Keyword(s):

Urban Heat Island ◽

Causal Discovery ◽

Heat Island ◽

Urban Heat ◽

Discovery Algorithms

Download Full-text

The lesson of causal discovery algorithms for quantum correlations: causal explanations of Bell-inequality violations require fine-tuning

New Journal of Physics ◽

10.1088/1367-2630/17/3/033002 ◽

2015 ◽

Vol 17 (3) ◽

pp. 033002 ◽

Cited By ~ 100

Author(s):

Christopher J Wood ◽

Robert W Spekkens

Keyword(s):

Bell Inequality ◽

Quantum Correlations ◽

Fine Tuning ◽

Causal Discovery ◽

Causal Explanations ◽

Discovery Algorithms

Download Full-text

Integrating Overlapping Datasets Using Bivariate Causal Discovery

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v34i04.5789 ◽

2020 ◽

Vol 34 (04) ◽

pp. 3781-3790

Author(s):

Anish Dhir ◽

Ciaran M. Lee

Keyword(s):

Conditional Independence ◽

Causal Structure ◽

Real Data ◽

Causal Discovery ◽

Causal Relations ◽

Multiple Datasets ◽

Single Dataset ◽

Markov Equivalence ◽

Causal Structures ◽

Discovery Algorithms

Causal knowledge is vital for effective reasoning in science, as causal relations, unlike correlations, allow one to reason about the outcomes of interventions. Algorithms that can discover causal relations from observational data are based on the assumption that all variables have been jointly measured in a single dataset. In many cases this assumption fails. Previous approaches to overcoming this shortcoming devised algorithms that returned all joint causal structures consistent with the conditional independence information contained in each individual dataset. But, as conditional independence tests only determine causal structure up to Markov equivalence, the number of consistent joint structures returned by these approaches can be quite large. The last decade has seen the development of elegant algorithms for discovering causal relations beyond conditional independence, which can distinguish among Markov equivalent structures. In this work we adapt and extend these so-called bivariate causal discovery algorithms to the problem of learning consistent causal structures from multiple datasets with overlapping variables belonging to the same generating process, providing a sound and complete algorithm that outperforms previous approaches on synthetic and real data.

Download Full-text