In Machine Learning/Data Mining, the class imbalanced problem occurs when there are many more instances of one class than the other. This problem occurs in a large number of real-world domains. Although a number of approaches have previously been proposed to tackle the problem, there still is little consensus about which approach is best or how to evaluate its performance. This talk will begin by reviewing some of the issues associated with class imbalances; it will then summarize the papers presented at the Workshop (these papers roughly fell in the following categories: comparisons of previously proposed schemes; new lights onto the issues; new approaches falling along the lines of previously proposed ones; novel approaches, novel problems). It will conclude with an overview of the main issues that came up during the (very lively!) discussion sessions.