### Bayesian methods

Bayesian methods rely on Bayes' theorem to make inferences according to the posterior distribution. These methods are particularly useful for obtaining insights about the data generation process.

Bayesian methods rely on Bayes' theorem to make inferences according to the posterior distribution. These methods are particularly useful for obtaining insights about the data generation process.

Data analysis typically involves data exploration, modeling, and presentation of the results. The posts in this series are concerned with comprehensive analyses that lead to new insights.

If your data are matched (e.g. paired), you should also treat it like this. Learn which statistical tests are appropriate for matched data here!

Hugo is a generator for static sites on the web. For example, this blog has been created with Hugo. Here, I discuss how Hugo can be used to create a blog, implement a commenting system, and much more.

Linear prediction models are among the most simple types of models. Despity their simplicity, these models can be very useful when certain conditions are met.

Parametric tests assume that the data follow a certain distribution. This page introduces how tests such as t-test, Chi-squared test, and ANOVA work in R.

When validating a model, it is crucial to select a suitable performance measure. This article discusses performance measures for regression (e.g. MSE, R squared) and classification (e.g. F1 statistic, AUC).

Plots are the most important tool for communicating results. However, choosing the right type of plot can be challenging. Learn here, how to choose between histograms, scatter plots, box plots, and so on.

R is extremely useful for data science because it is well-suited for applications in machine learning, data visualization, and statistics.

1/2
»