Occam's razor: from Ockham's via moderna to modern data science

Area of Science:

• Philosophy of Science
• Data Science
• Scientific Methodology

Background:

• The principle of parsimony, or Occam's razor, is a widely recognized heuristic in science.
• Despite its conceptual simplicity, rigorous formalization remains challenging.
• The 'big data' era intensifies the need for a robust understanding of parsimony.

Purpose of the Study:

• To review the historical significance of Occam's razor in William of Ockham's work.
• To survey contemporary developments and applications of parsimony in data science.
• To address the challenges of applying parsimony in the context of big data.

Main Methods:

• Historical literature review focusing on William of Ockham's philosophical and theological writings.
• Survey of modern data science literature on the application of parsimony principles.
• Analysis of the heuristic's utility and limitations in current scientific practice.

Main Results:

• Occam's razor has deep roots in philosophical thought, dating back to William of Ockham.
• Modern data science is increasingly employing parsimony principles for model selection and interpretation.
• Formalizing parsimony in big data contexts presents ongoing research challenges.

Conclusions:

• Parsimony remains a vital guiding principle in scientific inquiry.
• Bridging historical understanding with modern data science applications is crucial.
• Further research is needed to rigorously define and apply parsimony in the age of big data.