The paper presents a class binarization that combines fuzzy classifiers and coupled map lattices. First, a classification problem is divided into several two-class problems following an extended version of a fuzzy round robin class binarization scheme; next, a fuzzy classifier is generated using any machine learning technique for each two-class problem (we use evolution of fuzzy rules in this paper); finally, the generated fuzzy classifiers are integrated into a 2-dimensional coupled map lattice. The answer of the classifier to a sample is determined by the dynamics of the lattice when it is initialized with the answers given by each fuzzy classifier. Experiments are conducted with various publicly available data sets.