The multiclass classification problem is an important topic in the field of pattern recognition. It involves the task of classifying input instances into one of multiple classes. Since the class overlapping problem exists among multiple classes in most real-world problems, the multiclass classification task is much more complicated and challenging compared to the binary class problem.

The design of effective multiclass classifiers has been attracting the attention of a lot of researchers. Furthermore, the class imbalance problem hinders the study of many real-world applications. As a minority class is of a much smaller sample size compared with the majority class in most of the multiclass data sets, the prediction of traditional classifiers is biased towards the majority class, especially for high-dimensional data. To avoid great underestimation of the classification performance of the minority, novel methods are required for imbalanced data. The relationships among classes are not straightforward, and a lot of characteristics of samples within each class should be considered.

Therefore, the aim of this Special Issue is to attract high-quality papers from academics and industry researchers in data mining and machine learning and to present the most recent advanced methods and applications for effective research in the theory and application of the multiclass classification problem through scientific programming. Both original research and review articles are welcome.

Potential topics include but are not limited to the following:

• Methods and tools for multiclass classification
• Multiclass data analysis
• Feature selection/extraction methods for the multiclass problem
• Incremental/online learning techniques for the multiclass problem
• Industry and real-life applications
• Theory exploration on the multiclass problem in other fields