Basic Time Series Forecasting Techniques
An overview of some very simple forecasting models
Introduction
Forecasting is a wide domain with numerous applications in almost every industry. Due to this, the range of forecasting models is also very large with each model having its own pros and cons.
In this article, I want to go over some basic and simple forecasting models. Despite their simplicity, these models can offer good results in practice and provide a good basis to iterate from.
Average Forecast
The first model we will consider is the average forecast. This model simply assumes that all future values are equal to the mean of all the previous observations:
Where h is the future time-step we are forecasting for, T is the length of the time series, y_t is an observed value at time t and y_bar is the mean of the observed values. For this model we must have some past data available to compute the forecast.
