Information-Theoretic Causal Power

Korb, Kevin B.; Hope, Lucas R.; Nyberg, Erik

doi:10.1007/978-0-387-84816-7_10

Cited by 11 publications

(18 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…assigning a lower probability to a more general outcome than to one of the specific outcomes it includes (Jarvstad & Hahn, 2011), as well as more domain-specific versions of these general fallacies, such as the prosecutor's fallacy (Fenton & Neil, 2000) and the jury observation fallacy (Fenton & Neil, 2000). Similarly, the well-known trap of confusing correlation with causation repeatedly snares even good research scientists (some examples are discussed in Korb, Hope &Nyberg, 2009 andLawlor, Davey Smith &Ebrahim, 2004), which may well lead to overestimating the likely impact of policy actions. Common interrelationships between several variables can easily result in misunderstanding the significance of evidence, e.g.…”

Section: Psychology Of Probability and Causationmentioning

confidence: 99%

Improving the production and evaluation of structural models using a Delphi process

Bolger

Nyberg²,

Belton

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Bayes Nets (BNs) are extremely useful for causal and probabilistic modelling in many real-world applications, often built with information elicited from groups of domain experts. But their potential for reasoning and decision support has been limited by two major factors: the need for significant normative knowledge, and the lack of any validated methods or software supporting collaboration. Consequently, we have developed a web-based structured technique – Bayesian Argumentation via Delphi (BARD) – to enable groups of domain experts to receive minimal normative training and then collaborate effectively to produce high-quality BNs. BARD harnesses multiple perspectives on a problem, while minimising biases manifest in freely interacting groups, via a Delphi process: solutions are first produced individually, then shared, followed by an opportunity for individuals to revise their solutions. To test the hypothesis that BNs improve due to Delphi, we conducted an experiment whereby individuals with a little BN training and practice produced structural models using BARD for two Bayesian reasoning problems. Participants then received 6 other structural models for each problem, rated their quality on a 7-point scale, and revised their own models if they wished. Both top-rated and revised models were on average significantly better quality (scored against a gold-standard) than the initial models, with large and medium effect sizes. We conclude that Delphi – and BARD – improves the quality of BNs produced by groups. Further, although rating cannot create new models, rating seems quicker and easier than revision and yielded significantly better models – so, we suggest efficient BN amalgamation should include both.

show abstract

Section: Psychology Of Probability and Causationmentioning

confidence: 99%

Improving the production and evaluation of structural models using a Delphi process

Bolger

Nyberg²,

Belton

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…In addition, in this model, the arcs reflect the direction of causation. But many algorithms have been developed for learning causal of Bayesian Networks from data (Spirtes et al 2000;Neapolitan 2004;Korb et al 2009). For this reason, it should also possible to use the models based on various algorithms in line with the requirements and preferences of users.…”

Section: Figure 2: Dynamic Bayesian Network For the Avalanche Hazard Assessmentmentioning

confidence: 99%

Dinamik Çığ Tehlike Değerlendirmesi İçin Bayes Ağlarının CBS'ye Entegrasyonu: UKVA Perspektifi

Yılmaz

Öztürk

2018

Doğal Afetler Ve Çevre Dergisi

View full text Add to dashboard Cite

Natural hazard assessments are core to risk definition and early warning systems and play a fundamental role in the prevention of major damages. Traditional hazard identification methods are static. For this reason, new information and conditions cannot be easily included in the pre-defined hazard assessments. The Bayesian Networks can be used effectively for dynamic hazard identification. In this study, a methodology based on the Bayesian Networks model is presented for dynamic avalanche hazard assessment, in which changed and renewed data can be included in the system. In the proposed methodology, the integration of the Bayesian Networks and Geographical Information Systems (GIS) is modeled in the National Spatial Data Infrastructure (NSDI) perspective. In this structure, it is possible to combine and analyze the data obtained from different sources and factors for avalanche hazard can be dynamically updated with real-time updated data and temporal hazard mapping can be produced. The proposed methodology provides a generic structure and has an attribute making it applicable for dynamic mapping studies for other disasters.

show abstract

“…5 A treatment of the second aspect of Crick's ideas about information using algorithmic information measures is in preparation. 6 This measure has been independently proposed in neuroscience [23] and in the computational sciences [24]. For other related measures, see Ay & Polani [25] and Janzing et al [26].…”

Section: Endnotesmentioning

confidence: 99%

Genetic, epigenetic and exogenetic information in development and evolution

Griffiths¹

2017

Interface Focus.

View full text Add to dashboard Cite

The idea that development is the expression of information accumulated during evolution and that heredity is the transmission of this information is surprisingly hard to cash out in strict, scientific terms. This paper seeks to do so using the sense of information introduced by Francis Crick in his sequence hypothesis and central dogma of molecular biology. It focuses on Crick's idea of precise determination. This is analysed using an information-theoretic measure of causal specificity. This allows us to reconstruct some of Crick's claims about information in transcription and translation. Crick's approach to information has natural extensions to non-coding regions of DNA, to epigenetic marks, and to the genetic or environmental upstream causes of those epigenetic marks. Epigenetic information cannot be reduced to genetic information. The existence of biological information in epigenetic and exogenetic factors is relevant to evolution as well as to development.

show abstract

Information-Theoretic Causal Power

Cited by 11 publications

References 22 publications

Improving the production and evaluation of structural models using a Delphi process

Improving the production and evaluation of structural models using a Delphi process

Dinamik Çığ Tehlike Değerlendirmesi İçin Bayes Ağlarının CBS'ye Entegrasyonu: UKVA Perspektifi

Genetic, epigenetic and exogenetic information in development and evolution

Contact Info

Product

Resources

About