1. Forecasting models based on decomposition methods
2. Adding ETS and ARIMA filters
- 2.1 Exponential Smoothing
- 2.2 Identifying Stationarity
- 2.3 ARIMA models
- 2.4 Decomposition & ETS/ARIMA
3. Adding exogenous variables to the model
- 3.1 Linear Regression Models
- 3.2 Build dynamic regressions
- 3.3 Analyze the model's performance
- 3.4 Choose the best variables for the model
4. Forecasting at scale
- 4.1 Understanding the challenges and approaches when dealing with multiple time series
- 4.2 Applying hierarchical and reconciliation models
- 4.3 Efficient forecasting workflow
- 4.4 Evaluation of the global performance

Time Series Forecasting

Author

pbenavides

Modified

September 4, 2025

What is a time series?

Is this a time series?

If we focus solely on the regular plot, we wouldn’t have any time series. However, when we map each variable through time, we now have multiple time series: one for each country regarding life exp., GDP per capita, and population.

Financial Series

Stocks

Stocks, FX, … are all time series

Cryptos

Crypto currencies are also time series

Any variable that is measured through time is a time series.

What are forecasts?

flowchart LR
    A(There are two types of Data Scientists)
    A-->B(Those who can't predict the future)
    A-->C(Those who don't know that they can't predict the future)

No one, except for sorcerers and wizards, can predict the future.

What was Dr. Strange doing here?

Dr. Strange didn’t have the Time stone. He was using a high-tech gamer PC to run millions of simulations.

Can all variables be predicted with the same accuracy?

Eclipses

We can predict eclipses with complete certainty.

It’s not so easy to predict stock prices

Other variables can’t be predicted that easily. What does it depend on?

Electricity Demand

Forecasts

Beer Production Forecasts

Can you observe any strange patterns?

Which US Employment forecast works best?

Forecasting US Retail Employment using the Drift method

Forecasting US Retail Employment using the Seasonal Naive method

Forecasting US Retail Employment using ARIMA