In this work we will go through the analysis of **non-evenly spaced time series** data. We will make use of *"modern"* approaches to make these analysis, and compare their perfomance with less complex models that will act as our baseline.

We will create syntetic data of 3 random variables x1, x2 and x3, and adding some noise to the linear combination of some of the lags of these variables we will determine y, the response.

This way we can make sure that the function is not 100% predictable, the response depends on the predictors, and that there is a **time dependency** caused by the effect of previous **lags of the predictors** on the response.

Each of the models will start with a brief theoretical explanation and references to sources where more information can be found.

Then, documented code is provided and the conclusions of the results obtained.

We will go through the different models:

2 - Data Creation and Construction of Baseline Models