About this Research Topic
The challenge of deriving insights from the Internet of Things (IoT) has been recognized as one of the most exciting and key opportunities for both academia and industry. Advanced analysis of big data streams from sensors and devices is set to become a key area of data mining research as the number of applications requiring such processing increases. Dealing with imbalanced data that evolves over time, i.e., with concepts that drift or change completely, is one of the core issues in IoT stream mining.
This collection aims at discussing the problem of learning from IoT data streams generated by evolving non-stationary processes. It will focus on the advances of distributed algorithms and techniques, methods and tools that are dedicated to managing, exploit and interpret data streams in imbalance and non-stationary environments. It will focus on the problems of modelling, prediction, and classification based on learning from data streams.
The volume of data is rapidly increasing due to the development of the technology of information and communication. This data comes mostly in the form of streams. Learning from this ever-growing amount of data requires flexible learning models that self-adapt over time. In addition, these models must consider many constraints: (pseudo) real-time processing, high-velocity, and dynamic multi-form change such as concept drift and novelty. Consequently, learning from streams of evolving and unbounded data requires developing new algorithms and methods able to learn under the following constraints:
· random access to observations is not feasible or it has high costs,
· memory is small with respect to the size of data,
· data distribution or phenomena generating the data may evolve over time, which is known as concept drift and
· the number of classes may evolve overtime. Therefore, efficient data streams processing requires drivers and learning techniques:
· Incremental learning in order to integrate the information carried by each new arriving data;
· Decremental learning in order to forget or unlearn the data samples which are no more useful;
· Novelty detection to learn new concepts.
It is worthwhile to emphasize that streams are very often generated by distributed sources, especially with the advent of Internet of Things. Scalable and decentralized learning algorithms are potentially more suitable and efficient.
This collection provides a forum to discuss important research questions and practical challenges in IoT big data stream mining. We welcome papers; describing and evaluating new (distributed) streaming learning algorithms, controversial issues related to standards, methods, tasks and evaluation, open problems, visualization, applications, and comparisons of competing approaches are strongly encouraged. Representation of alternative viewpoints and discussions are also strongly encouraged. We invite submission of papers describing innovative research on all aspects of IoT big data stream mining. Papers emphasizing theoretical foundations, algorithms, systems, distributed and parallel methods, federate learning, applications, language issues, data storage, access, and architecture are
Keywords: Data Streams, Distributed Data Mining, Spatio-temporal Data Mining, Internet of Things, Big Data
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.