liveBook · Manning

Frequentist paradigm

As explained in Math and Architectures of Deep Learning

A probability interpretation that defines probability as relative frequency within a population or repeated trials under a specified condition.

The frequentist paradigm defines probability by repeatedly observing a phenomenon and measuring how often an event occurs. For a specified population, divide favorable outcomes by total outcomes. section 5.1 section 6.0

The counting interpretation supports several probability constructions:

  • For a discrete event, count members satisfying the criterion and divide by the total population; the result cannot exceed 1. section 5.1
  • For conditional probability, restrict the universe to the population satisfying the given condition, then count the target event within that restricted population. section 6.1.2
  • For continuous variables, refine value buckets toward infinitesimal width; the limiting probabilities are represented by a probability density function. section 5.6

A population count

If Statsville has 100,000 adults and 25,000 are at least 6 feet tall, the frequentist probability of selecting an adult meeting that criterion is 25,000 / 100,000 = 0.25. For a conditional question, such as weight over 90 kg given height between 160 and 183 cm, use only the 48,000 people in that height range as the population. section 5.1 section 6.1.2

Equation 5.1

Use the equation view to inspect the favorable-outcomes-over-total-population formulation. section 5.1

How Math and Architectures of Deep Learning covers it

How deep it goes
Develops counting, conditional restriction, and continuous-density limits, but does not detail estimation procedures or computational implementation.
Examples
Uses Statsville population counts, height and weight categories, conditional-probability tables, and infinitesimal value buckets.
What you can do after
Apply population counts to calculate discrete and conditional probabilities and interpret continuous probabilities through density.
How it connects
Links probability to conditional probability, probability density functions, Bayesian statistics, and machine learning.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Math and Architectures of Deep Learning. The section links point to the parts of the book it draws on.