Causality, Probability, and Time

Abstract

Causality is a key part of many fields and facets of life, from finding the relationship between diet and disease to discovering the reason for a particular stock market crash. Despite centuries of work in philosophy and decades of computational research, automated inference and explanation remains an open problem. In particular, the timing and complexity of relationships has been largely ignored even though this information is critically important for prediction, explanation and intervention. However, given the growing availability of large observational datasets including those from electronic health records and social networks, it is a practical necessity. This book presents a new approach to inference (finding relationships from a set of data) and explanation (assessing why a particular event occurred), addressing both the timing and complexity of relationships. The practical use of the method developed is illustrated through theoretical and experimental case studies, demonstrating its feasibility and success.

Tetraeffective causes, mortacauses, and vitacauses of mortality and survivorship

“…Scientific research on causality remains problematic and challenging [19,[122][123][124][125][126], scientific research on causes of mortality and survivorship remains particularly problematic and challenging [127][128][129][130][131][132][133][134][135], and mortality and survivorship and their interrelationships are particularly prone to elicit errors and biases [136][137][138][139][140][141]. The investigation here shows that consideration of tetraeffective causes of mortality and survivorship usefully elucidatesand deepens the consideration of, and expands the scope of scientific research oncauses of mortality and survivorship.…”

“…Thus considerations of tetraeffective causes of mortality and survivorshipand the methods and procedures that are employed hereaddress diverse problems and meet diverse challenges. However, other diverse problems and challenges remain, and other diverse problems and challenges comeand will continue to comeinto focus; many of these problems and challenges require further research with diverse methods and procedures [19,[122][123][124][125][126][127][128][129][130][131][132][133][134][135].…”

Temporal Causality Analysis of Sentiment Change in a Cancer Survivor Network

“…Suppes [25] offers a method for testing whether a cause is spurious in the restricted setting where there are only two possible causes. Kleinberg [24] argued for a more stringent criterion, for a prima facie cause to be considered a genuine cause and introduced a method for assessing the causal significance of a potential cause of an effect which can be used to identify a genuine cause of an event from among a set of its potential causes.…”

“…However, both dynamic Bayesian networks and Granger causality lack the expressivity needed to represent and reason about the temporal properties of the underlying system, e.g., “the first positive reply to a post expressing a negative sentiment, with at least 80% probability, results in a change in sentiment within 5 hours”. Against this background, we introduce a novel approach that leverages the machinery of temporal causality developed in [24] to uncover the temporal causality of the dynamics of sentiment change (on the part of the thread originators) in OHCs. This approach explicitly captures the temporality of the relationship between cause and effect.…”

Nonparametric causal inference for bivariate time series

“…For example, there is no current technology that can control the interaction between the solar wind and the magnetic field measured at the surface of Earth, so space weather studies rely on data collected without performing controlled experiments. As a result, causal inference with observational data sets from such systems is difficult and the need to identify causal relationships given the weakness of correlation in doing so has lead to the development of several different time series causality tools [1][2][3][4][5].…”