Electroencephalography (EEG) is a widely used neuroimaging technique that records the electrical activity of the brain. EEG analysis provides valuable insights into brain dynamics and understanding of neural processes. As EEG data analysis relies heavily on signal processing and statistical analysis, it is crucial to have a robust framework for analyzing EEG data that produces reliable results. One very useful framework for EEG data analysis is the use of algorithmic complexity measures. Algorithmic complexity is a measure of the complexity of a given sequence of data such as the EEG waveform. It provides a way to quantify the amount of randomness and predictability within EEG data. Along with traditional complexity measures like Sample Entropy, Hurst Exponent, Multiscale Entropy, etc., there is a less-known approach involving Kolmogorov-Chaitin algorithmic complexity, which is a mathematical approach used for measuring the complexity of a string of information. It is based on the idea that a complex string of information cannot be compressed or represented by a simpler algorithm. The advantages of using Kolmogorov-Chaitin complexity include its objectivity, non-linearity, ability to capture content and robustness. This paper presents the basics of the later approach and shows how it can be used for machine learning on EEG data.