I took the Solve It with Code course from answer.ai, and Advent of Code is used as practice. I always thought my Python skills were quite mediocre; I always used to in a “just get things done” style, so this is also a chance to improve my Python. For this reason I want to mostly stick with the standard packages and pythonic styles, and only use extra packages when it’s too cumbersome otherwise. ...
intro Bayesian modeling of global trend, local variation, seasonality, and heterogeneity in time series In this post we’ll implement a time series framework based on a flexible smoothed exponential process. The framework can be used to model global trend, local variation, seasonality and other features that are essential for flexible time series modeling. This post also covers several other features under this general framework, notably: modeling over-dispersed observations with Student-t distribution explicit and extensive modeling of the variation to capture data heteroscedasticity adding regression components when necessary Some other general but important features include: extensive usage of visualisation for model criticism prior and posterior simulation for model inspection and a unified prediction framework. Lastly but most importantly, apart from the specific features and techniques, the post gives a general procedure for Bayesian model building and criticism, which should be useful for building any kind of model for any kind of data. ...