AI/ML Overview

Bayesian and Frequentist Approaches

Bayesian = Probability of Hypothesis + Probability of Data

Frequentist = Probability of Data

Understanding

Approach:

• Apply a mechanistic model

  • Multivariate Regression

  • Differential Equations

• Goal: Understand the relative weighting of parts and how they interact.

• Result: Estimation of how a natural phenomena works.

Example:

Research question: How does insolation drive sea surface temperature?

• Frequentist Approach: Compare measurements of insolation and sea surface temperature and establish a coefficient of relation based on observational data.

• Bayesian Approach: Based on previous knowledge and studies of this relationship, update the distribution of the relation with new observational data to produce an updated distribution of the coefficient of relationship.

Prediction

Approach:

• Apply a model and estimate it’s predictive performance (not necessarily it’s goodness of fit)

  • Pretty much any model

  • Including ‘Black Box’ Models

• Goal: Reproducibly produce accurate and precise estimates of desired outcome given inputs.

• Result: A tool to map inputs to outputs with known performance.

Example:

Research question: How does insolation drive sea surface temperature?

• Build models on subsets of data that link insolation with sea surface temperature.

• Measure the accuracy and precision of these models in their linking ability.

• Keep the better models and return the model as a mapping function with its performance metrics