Artificial intelligence is a challenging new discipline of science that tries to combine cognition with computers and computational methodology. A complex system on the other hand is a physical, technological, or biological system that is generally difficult to formulate precisely or solve mathematically and may lead to interesting and sometimes puzzling behaviors. Merging artificial intelligence with complex systems may lead to practical solutions for problems that are both important and generally difficult to handle with other means. We give a small tour of ideas in complex systems and also initiate the study of machine learning. The end result of a machine learning process is a model that has learned to fit data and can determine properties of other data. The model itself is a function or more generally an algorithm that may be applied to datasets. Learning can take place in several ways, the more usual one being supervised, although unsupervised or semisupervised and reinforcement being viable alternatives. Data are the central object in this process, and they are used in order to generate the correct model and also test it. Data is split into a training set and a test set; the former trains the model and determines values of its parameters that are appropriate. The test set part of the data is not used during training, but it is applied to the trained model in order to test how well it works with controlled but unknown data. In order for the model to generalize well, i.e., to be able to work well with new and unknown data, overfitting must be avoided.