In the real world concepts are often not stable but change with time. Typical examples of this are weather prediction rules and customers' preferences. The underlying data distribution may change as well. Often these changes make the model built on old data inconsistent with the new data, and regular updating of the model is necessary. This problem, known as concept drift, complicates the task of learning a model from data and requires special approaches, different from commonly used techniques, which treat arriving instances as equally important contributors to the final concept. This paper considers different types of concept drift, peculiarities of the problem, and gives a critical review of existing approaches to the problem.

A surprisingly large number of research disciplines have contributed towards the development of knowledge on lazy problem solving, which is characterized by its storage of ground cases and its demand driven response to queries. Case-based reasoning (CBR) is an alternative, increasingly popular approach for designing expert systems that implements this approach. This paper lists pointers to some contributions in some related disciplines that offer insights for CBR research. We then outline a small number of Navy applications based on this approach that demonstrate its breadth of applicability. Finally, we list a few successful and failed attempts to apply CBR, and list some predictions on the future roles of CBR in applications.

. Case-based problem-solving systems reason and learn from experiences, building up case libraries of problems and solutions to guide future reasoning. The expected bene#ts of this learning process depend on twotypes of regularity: #1# problem-solution regularity, the relationship between problem-to-problem and solution-to-solution similaritymeasures that assures that solutions to similar prior problems are a useful starting point for solving similar current problems, and #2# problemdistribution regularity, the relationship between old and new problems that assures that the case library will contain cases similar to the new problems it encounters. Unfortunately, these types of regularity are not assured. Even in contexts for which initial regularityissu#cient, problems may arise if a system's users, tasks, or external environmentchange over time. This paper de#nes criteria for assessing the twotypes of regularity, discusses how the de#nitions may be used to assess the need...

This paper presents local methods for modeling and control of discrete-time unknown nonlinear dynamical systems, when only a limited amount of input-output data is available. We propose the adoption of lazy learning, a memory-based technique for local modeling. The modeling procedure uses a query-based approach to select the best model configuration by assessing and comparing different alternatives. A new recursive technique for local model identification and validation is presented, together with an enhanced statistical method for model selection. Also, three methods to design controllers based on the local linearization provided by the lazy learning algorithm are described. In the first method the lazy technique returns the forward and inverse models of the system which are used to compute the control action to take. The second is an indirect method inspired to self-tuning regulators where recursive least squares estimation is replaced by a local approximator. The third method combin...

Spam filtering is a particularly challenging machine learning task as the data distribution and concept being learned changes over time. It exhibits a particularly awkward form of concept drift as the change is driven by spammers wishing to circumvent spam filters. In this paper we show that lazy learning techniques are appropriate for such dynamically changing contexts. We present a case-based system for spam filtering that can learn dynamically. We evaluate its performance as the case-base is updated with new cases. We also explore the benefit of periodically redoing the feature selection process to bring new features into play. Our evaluation shows that these two levels of model update are effective in tracking concept drift.

Abstract. In this paper, we compare case-based spam filters, focusing on their resilience to concept drift. In particular, we evaluate how to track concept drift using a case-based spam filter that uses a featurefree distance measure based on text compression. In our experiments, we compare two ways to normalise such a distance measure, finding that the one proposed in [1] performs better. We show that a policy as simple as retaining misclassified examples has a hugely beneficial effect on handling concept drift in spam but, on its own, it results in the case base growing by over 30%. We then compare two different retention policies and two different forgetting policies (one a form of instance selection, the other a form of instance weighting) and find that they perform roughly as well as each other while keeping the case base size constant. Finally, we compare a feature-based textual case-based spam filter with our feature-free approach. In the face of concept drift, the feature-based approach requires the case base to be rebuilt periodically so that we can select a new feature set that better predicts the target concept. We find feature-free approaches to have lower error rates than their feature-based equivalents. 1

Abstract: Mining data streams is a challenging task that requires online systems based on incremental learning approaches. This paper describes a classification system based on decision rules that may store up–to–date border examples to avoid unnecessary revisions when virtual drifts are present in data. Consistent rules classify new test examples by covering and inconsistent rules classify them by distance as the nearest neighbour algorithm. In addition, the system provides an implicit forgetting heuristic so that positive and negative examples are removed from a rule when they are not near one another.

In the field of system identification and control a mismatch exists between the available theoretical tools and most of the problems encountered in practice. On the one hand, researchers developed plenty of theoretical analysis and methods concerning linear systems; on the other hand practitioners are often confronted with the apparent nonlinearity of the real world. Although several nonlinear identification and control techniques have been proposed in the last decades, these still appear to be less robust and reliable than their linear counterparts. An appealing approach to bridge the existing gap consists in decomposing a complex nonlinear control problem in a number of simpler linear problems, each associated with a restricted operating region. This paper will review a number of divide-and-conquer techniques proposed in the nonlinear control literature and more recently in the machine learning community to address the problem of linearly controlling a nonlinear system. Two families of divide-and-conquer approaches will be taken into consideration: analytical approaches which require the knowledge of the system dynamics and learning approaches which rely on powerful approximators to estimate a model from input/output data.