The Automated Laplacean Demon: How ML Challenges Our Views on Prediction and Explanation

Srećković, Sanja; Berber, Andrea; Filipović, Nenad R.

doi:10.1007/s11023-021-09575-6

Cited by 11 publications

(4 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This concept raises reasonable questions about the nature of free will, the possibility of predicting the future, and the limitations of our knowledge of the world. LaPlace's demon remains an interesting and controversial issue that makes us think about the nature of reality, free will, and the limits of human knowledge (Srećković et al, 2021).…”

Section: Resultsmentioning

confidence: 99%

Free will: Philosophical views on the nature of human choice and limitations

Adli,

Abykeeva-Sultanalieva

2024

European Chronicle

View full text Add to dashboard Cite

The need to explore free will arises from continuous reflection and discussion about the nature of human existence and the ability of humans to make free decisions. This need is compounded by ambiguity in views and approaches to this concept and its meaning for various aspects of life. The examination of free will becomes relevant due to the need to understand the causes of actions, moral responsibility, and the possibilities of human influence on the world. This study was devoted to the disclosure of key aspects of free will through the analysis of historical and modern theories. In the course of the study, a literary analysis was involved, which included work with papers on related subjects. The conceptual analysis allowed defining and understanding the main concepts and terms of the subject under study. Comparative analysis was applied to the papers of other researchers who worked on the same or similar subjects to identify similarities and differences and determine prospects for future research in this area. The study shows that there are two main directions in this discussion: determinists believe that all events, including human actions, are predetermined, and compatibilists state that free will is compatible with determinism. The study provides a detailed analysis of deterministic and compatibilist concepts of free will. Arguments for and against free will, which indicate the influence of external factors on human choice, are considered. Through the analysis of historical and modern theories, the examination of various concepts and argumentation, this study identifies the main approaches to the interpretation of the concept of free will in the context of philosophical thought. The main results indicate the complexity and versatility of the concept of free will and the importance of considering various factors that influence human choice

show abstract

Section: Resultsmentioning

confidence: 99%

Free will: Philosophical views on the nature of human choice and limitations

Adli,

Abykeeva-Sultanalieva

2024

European Chronicle

View full text Add to dashboard Cite

show abstract

“…Creel points out that to the extent that explanation requires insight into underlying mechanism, opacity will be an obstacle to the explanatory goals of science (Creel, 2020). Srećković et al (2022) imagine that machine learning is likely to lead to the emergence of two separate strands of scientific activity, one "purely predictive and detached from any explanatory efforts", and one that is "faithful to the anthropocentric research focused on the search for explanation" (Srećković et al, 2022, 179). Finally, Boge (2022) argues that deep neural nets are purely instrumental models that do not deliver explanations and stand in the way of proper conceptualizations of the target system.…”

Section: Machine Learning and The End Of Theorymentioning

confidence: 99%

The Explanatory Role of Machine Learning in Molecular Biology

Gross

2024

Erkenn

View full text Add to dashboard Cite

The philosophical debate around the impact of machine learning in science is often framed in terms of a choice between AI and classical methods as mutually exclusive alternatives involving difficult epistemological trade-offs. A common worry regarding machine learning methods specifically is that they lead to opaque models that make predictions but do not lead to explanation or understanding. Focusing on the field of molecular biology, I argue that in practice machine learning is often used with explanatory aims. More specifically, I argue that machine learning can be tightly integrated with other, more traditional, research methods and in a clear sense can contribute to insight into the causal processes underlying phenomena of interest to biologists. One could even say that machine learning is not the end of theory in important areas of biology, as has been argued, but rather a new beginning. I support these claims with a detailed discussion of a case study involving gene regulation by microRNAs.

show abstract

“…Traditionally, the ability to generate accurate predictions has been viewed as the hall mark of successful scientific explanations (for overviews see Douglas, 2009;Srećković et al, 2022). In a general sense, this is true; but more faithful explanations are guaranteed to yield more accurate predictions only under ideal circumstances, in which measurement and sampling error are absent or practically negligible.…”

Section: Prediction ≠ Explanationmentioning

confidence: 99%

“…Or, it is entirely possible for the same variable to improve the performance of a predictive model, but seriously distort the results of an explanatory model (for example because it acts as a collider for the effect of interest; see Elwert & Winship, 2014;Rohrer, 2018). Correctly representing the causal relations between variables is essential for scientific explanation, but unnecessary and usually irrelevant for prediction (see Srećković et al, 2022). For these reasons, predictive and explanatory accuracy may come at each other's expense regardless of the complexity and number of parameters of the models employed.…”

Section: Tradeoffs Between Prediction and Explanation In Machine Lear...mentioning

confidence: 99%

The prediction-explanation fallacy: A pervasive problem in scientific applications of machine learning

Del Giudice

2024

Methodology

View full text Add to dashboard Cite

I highlight a problem that has become ubiquitous in scientific applications of machine learning and can lead to seriously distorted inferences. I call it the Prediction-Explanation Fallacy. The fallacy occurs when researchers use prediction-optimized models for explanatory purposes, without considering the relevant tradeoffs. This is a problem for at least two reasons. First, prediction-optimized models are often deliberately biased and unrealistic in order to prevent overfitting. In other cases, they have an exceedingly complex structure that is hard or impossible to interpret. Second, different predictive models trained on the same or similar data can be biased in different ways, so that they may predict equally well but suggest conflicting explanations. Here I introduce the tradeoffs between prediction and explanation in a non-technical fashion, present illustrative examples from neuroscience, and end by discussing some mitigating factors and methods that can be used to limit the problem.

show abstract

The Automated Laplacean Demon: How ML Challenges Our Views on Prediction and Explanation

Cited by 11 publications

References 70 publications

Free will: Philosophical views on the nature of human choice and limitations

Free will: Philosophical views on the nature of human choice and limitations

The Explanatory Role of Machine Learning in Molecular Biology

The prediction-explanation fallacy: A pervasive problem in scientific applications of machine learning

Contact Info

Product

Resources

About