Abstract. An important component of many data mining projects is finding a good classification algorithm, a process that requires very careful thought about experimental design. If not done very carefully, comparative studies of classification and other types of algorithms can easily result in statistically invalid conclusions. This is especially true when one is using data mining techniques to analyze very large databases, which inevitably contain some statistically unlikely data. This paper describes several phenomena that can, if ignored, invalidate an experimental comparison. These phenomena and the conclusions that follow apply not only to classification, but to computational experiments in almost any aspect of data mining. The paper also discusses why comparative analysis is more important in evaluating some types of algorithms than for others, and provides some suggestions about how to avoid the pitfalls suffered by many experimental studies.

We are interested in the development of activities in situated, embodied agents such as mobile robots. Central to our theory of development is means-ends analysis planning, and as such, we must rely on operator models that can express the eects of a robot's action in a dynamic, partially-observable environment. This paper presents a two-step process which employs clustering and decision tree induction to perform unsupervised learning of operator models from simple interactions between an agent and its environment. We report our ndings with an implementation of this system on a Pioneer-1 mobile robot.