https://machinelearningmastery.com/time-series-forecasting/

<aside> 💡 A time series dataset has an explicit order dependences between observations, a time dimension.

</aside>

What this means in simple terms is that each observation has a corresponding value of time.

Here we are attempting to predict the future. An example would be predicting expenses for 26-03-2024 based on prior information.

Note that it is not always necessary or advised to understand the dataset (descriptive modelling) as this can be quite demanding technically and in terms of time required. We are more interested in predicting future values.

Forecasting

<aside> 💡 Forecasting involves taking models fit on historical data and using them to predict future observations.

</aside>

Here the future is unknown to us and must be estimated from what is already known to us (historical data).

The skill of a time series forecasting model is determined by its performance at predicting the future. This is often at the expense of being able to explain why a specific prediction was made, confidence intervals and even better understanding the underlying causes behind the problem.

Components of a Time Series

1. Level. The baseline value for the series if it were a straight line.
2. Trend. The optional and often linear increasing or decreasing behavior of the series over time.
3. Seasonality. The optional repeating patterns or cycles of behavior over time.
4. Noise. The optional variability in the observations that cannot be explained by the model.

Practical Example

This tutorial goes how one can fit a model to predict Bitcoin prices. A variation of the Autoregressive Moving Average (ARMA) is being used here.

https://builtin.com/data-science/time-series-forecasting-python

Key Takeways:

• Hyper-parameters are very important
  • Invest in a hyper-parameter tuning approach to get better predictions
  • This is because data can have seasonal trends that need to be accounted
• Have a meaningful test and train split
  • This becomes very important for evaluation
• Try different methods, models and evaluate the prediction.
  • RMSE is one way of evaluating the accuracy