The interest in developing machine learning (ML) algorithms for streaming data research has increased over the years. Complex problems provide data with new characteristics in comparison to the previous decade. Nowadays, data sources generate high dimensional information in large volumes and velocity. The generated data are becoming increasingly ubiquitous. To extract useful information is a real challenge. The provided data does not have a homogeneous structure, and it can also be redundant, noisy, and incomplete. These characteristics have motivated the development of several machine learning algorithms for data streams analysis. Streaming data research has mainly focused on developing accurate decision models with the ability to learn and forget concepts incrementally. The advances in streaming data research have been useful in areas such as clustering, temporal learning, anomaly detection, semi-supervised learning, novel class detection, and feature selection. ML research for data streams often appears on the big data (BD) analysis domain, related to tackling issues concerning to velocity and volume of the BD.