Incremental learning from data streams is increasingly attracting research focus due to many real streaming problems (such as learning from transactions, sensors or other sequential observations) that require processing and forecasting in the real time. In this paper we deal with two issues related to incremental learningprediction accuracy and prediction explanationand demonstrate their applicability on several streaming problems for predicting electricity load in the future. For improving prediction accuracy we propose and evaluate the use of two reliability estimators that allow us to estimate prediction error and correct predictions. For improving interpretability of the incremental model and its predictions we propose an adaptation of the existing prediction explanation methodology, which was originally developed for batch learning from stationary data. The explanation methodology is combined with a state-of-the-art concept drift detector and a visualization technique to enhance the explanation in dynamic streaming settings. The results show that the proposed approaches can improve prediction accuracy and allow transparent insight into the modeled concept.

The temporal dimension that is ever more prevalent in data makes the data stream mining (incremental learning) an important field of machine learning. In addition to accurate predictions, explanations of models and examples are a crucial component as they provide insight into model's decision and lessen its black box nature, thus increasing the user's trust. Proper visual representation of data is also very relevant to user's understanding -visualization is often utilised in machine learning since it shifts the balance between perception and cognition to take fuller advantage of the brain's abilities. In this paper we review visualisation in incremental setting and devise an improved version of an existing visualisation of explanations of incremental models. We discuss the detection of concept drift in data streams and experiment with a novel detection method that uses the stream of model's explanations to determine the places of change in the data domain.