The Living Thing / Notebooks :

Time series inference in discrete time

for dwellers in quantised synchronous realities

GOTO ARIMA, if you want to extend basic linear statistics to dependent series. GOTO state filter inference, if you want to use Bayes’ rule to update an estimate in time, maybe even non-linearly. GOTO High frequency time series, if you are given a set interval and have to decide how densely to sample it, because you want to do something quantish sounding for a financial institution.

It’s probably worth having a page for high dimensional time series.

Here’s some miscellaneous generic stuff:

prophet (R/Python/Stan)

is a procedure for forecasting time series data. It is based on an additive model where non-linear trends are fit with yearly and weekly seasonality, plus holidays. It works best with daily periodicity data with at least one year of historical data. Prophet is robust to missing data, shifts in the trend, and large outliers.


Automatic Smoothing for Attention Prioritization in Time Series

ASAP automatically smooths time series plots to remove short-term noise while retaining large-scale deviations.