A review of frequentist regression using lm(), an introduction to Bayesian regression using stan_glm(), and a comparison of the respective outputs.

Regressione lineare non bayesiana

Esplorare i dati

Adattare una regressione lineare frequentista

Regressione lineare bayesiana

Adattare una regressione lineare bayesiana

Criteri di convergenza

Valutare la convergenza del modello

Confrontare i metodi frequentisti e bayesiani

Differenza tra frequentisti e bayesiani

Creare intervalli credibili

Introduction to Bayesian Linear Models

Learn how to modify your Bayesian model including changing the number and length of chains, changing prior distributions, and adding predictors.

What's in a Bayesian Model?

Altering chains

Do I have enough iterations?

Prior distributions

Determine Prior Distributions

Calculate Adjusted Scales

Unadjusted Priors

User Specified Priors

Changing Priors

Specifying informative priors

Consequences of informative priors

Altering the estimation process

Altering the Estimation

Modifying a Bayesian Model

In this chapter, we'll learn how to determine if our estimated model fits our data and how to compare competing models.

Using the R Squared statistic

Calculating Frequentist R-squared

R-squared for a Bayesian Model

Posterior predictive model checks

Predicted score distributions

Distributions for a single observation

Model fit with posterior predictive model checks

R-squared Posterior

Posterior Predictive Testing

Bayesian model comparisons

Calculating the LOO estimate

Comparing models

Assessing Model Fit

In this chapter, we'll learn how to use the estimated model to create visualizations of your model and make predictions for new data.

Visualizing a Bayesian model

Plotting a Bayesian model

Plotting Model Uncertainty

Making predictions

Popularity for Observed Songs

Popularity for New Songs

Visualizing predictions

Format prediction posteriors

Visualize New Predictions

Conclusion

Presenting and Using a Bayesian Regression

Spotify dataset

La stima bayesiana offre un’alternativa flessibile alle tecniche di modellazione in cui le inferenze dipendono dai p-value. In questo corso imparerai a stimare modelli di regressione lineare usando metodi bayesiani e il pacchetto rstanarm. Sarai introdotto alle distribuzioni a priori, al controllo predittivo a posteriori del modello e al confronto tra modelli nel quadro bayesiano. Imparerai anche a usare il modello stimato per fare previsioni su nuovi dati.

Bayesian Modeling with RJAGS

Introduction to Data Visualization with ggplot2

Intermediate Regression in R

Scopri come utilizzare i metodi di stima bayesiana per migliorare le inferenze sui modelli di regressione.

Modellazione di regressione bayesiana con rstanarm

Scopri come usare i metodi di stima bayesiani per fare previsioni più accurate sui modelli di regressione lineare.

Scienziato dell'apprendimento automatico in R

Psychometrician, ATLAS, University of Kansas

Comparing Bayesian and Frequentist Approaches

The same parameters!

Frequentist vs. Bayesian

Evaluating Bayesian parameters

Creating credible intervals

Confidence vs. Credible intervals

Let's practice!

Introduction to Bayesian Linear Models - Confrontare i metodi frequentisti e bayesiani

Confrontare i metodi frequentisti e bayesiani

Create Your Free Account