What is Statistics? (Michael I. Jordan)  AI Podcast Clips  Summary and Q&A
TL;DR
Statistics is a discipline that allows for making informed decisions and inferences based on data, with the field originating from the study of inverse probability and evolving to incorporate decision theory.
Key Insights
 The field of statistics is a combination of math, science, and technology, allowing for inferences and decisionmaking with some level of certainty.
 Statistics originally emerged as "inverse probability," with a focus on understanding outcomes based on the underlying mechanism.
 Laplace played a crucial role in formalizing the field of statistics, leading to its current name.
 Statistics is closely related to game theory, decision theory, computer science, control, and economics, with decision theory being a fundamental aspect.
 Statistics involves considering loss functions, probability models, and the risks associated with different decisions.
 There are two main perspectives in statistics: Bayesian and frequentist.
 Bayesian statistics focuses on the unknown parameter, while frequentist statistics focuses on the random data.
 Empirical Bayes is an approach that combines the Bayesian framework with practical estimations, providing a reliable solution under certain assumptions.
 False discovery rate is a valuable concept in statistics that considers the proportion of false discoveries among all the discoveries made, providing a different criterion for accuracy.
Transcript
an absurd question but what is statistics so the here it's a little bit it's somewhere between math and science and technology it's somewhere in that convex hull so it's some principles that allow you to make inferences that have got some reason to be believed and also principle allow you to make decisions where you can have some reason to believe ... Read More
Questions & Answers
Q: What is the historical origin of the field of statistics?
The field of statistics traces its roots back 250 years when it was known as inverse probability, focusing on using outcomes to infer underlying mechanisms. It was originally developed to explain gambling situations but evolved to include decisionmaking and data analysis.
Q: What is the role of decision theory in statistics?
Decision theory is a crucial component of statistics, as it involves defining loss functions and addressing the question of what outcomes or decisions are desired. It provides a framework for considering uncertainty, evaluating risk, and guiding decisionmaking based on probability models and potential errors.
Q: What are Bayesian and frequentist approaches in statistics?
Bayesian and frequentist approaches are two different perspectives within statistics. Bayesian thinking focuses on inferring parameters based on data and using prior probabilities, while frequentist thinking focuses on analyzing the distribution of data and making inferences based on observed outcomes. Both approaches have their strengths and can provide valuable insights depending on the context.
Q: What is the false discovery rate in statistics?
The false discovery rate is a concept in statistics that deals with multiple hypothesis testing. It refers to the proportion of false discoveries among the total discoveries made in a set of tests. The false discovery rate criterion aims to keep this proportion small, ensuring the accuracy and reliability of findings. It can be approached from both frequentist and Bayesian perspectives, with empirical Bayesian methods offering reasonable estimates for the false discovery rate.
Summary & Key Takeaways

Statistics is a formal discipline that combines math, science, and technology to make inferences and decisions based on data.

The history of statistics dates back 250 years when it was known as inverse probability, focusing on using outcomes to infer underlying mechanisms.

Decision theory is a fundamental component of statistics, with the field branching out into Bayesian and frequentist approaches.