Abstract
Data stream is a challenging research topic in which data can continuously arrive with a probability distribution that may change over time. Depending on the changes in the data distribution, different phenomena can occur, for example, a concept drift. A concept drift occurs when the concepts associated with a dataset change when new data arrive. This paper proposes a new method based on k-Nearest Neighbors that implements a sliding window requiring less instances stored for training than existing methods. For such, a clustering approach is used to summarize data by placing labeled instances considered similar in the same cluster. Besides, instances close to the uncertainty border of existing classes are also stored, in a sliding window, to adapt the model to concept drift. The proposed method is experimentally compared with state-of-the-art classifiers from the data stream literature, regarding accuracy and processing time. According to the experimental results, the proposed method has better accuracy and less time consumption when fewer information about the concepts are stored in a single sliding window. © 2018, Springer Nature Switzerland AG.

Abstract
Human Activity Recognition has been primarily investigated as a machine learning classification task forcing it to handle with two main limitations. First, it must assume that the testing data has an equal distribution with the training sample. However, the inherent structure of an activity recognition systems is fertile in distribution changes over time, for instance, a specific person can perform physical activities differently from others, and even sensors are prone to misfunction. Secondly, to model the pattern of activities carried out by each user, a significant amount of data is needed. This is impractical especially in the actual era of Big Data with effortless access to public repositories. In order to deal with these problems, this paper investigates the use of Transfer Learning, specifically Unsupervised Domain Adaptation, within human activity recognition systems. The yielded experiment results reveal a useful transfer of knowledge and more importantly the convenience of transfer learning within human activity recognition. Apart from the delineated experiments, our work also contributes to the field of transfer learning in general through an exhaustive survey on transfer learning for human activity recognition based on wearables. © 2018, Springer Nature Switzerland AG.

Abstract
We present a method for improving the prediction accuracy using multiple predictive algorithms. Several techniques have been developed to tackle this issue such as bagging, boosting and stacking. In contrary to the first two that, usually, generate homogeneous ensembles of classifiers, stacking techniques have demonstrated success using heterogeneous ensembles. In our method, we adopt the stacking mechanism. Several models are generated using different learning algorithms. Forward stepwise selection is implemented to link each instance to its appropriate learning model. Experiments with three datasets benchmarked with four learning schemes show that this novel method improves prediction accuracy and can serve as a bridge to transfer knowledge between tasks given the same feature space but different data distributions. © 2018, Springer Nature Switzerland AG.

Abstract

Abstract
An efficient traffic-network is an essential demand for any smart city. Usually, city traffic forms a huge network with millions of locations and trips. Traffic flow prediction using such large data is a classical problem in intelligent transportation system (ITS). Many existing models such as ARIMA, SVR, ANN etc, are deployed to retrieve important characteristics of traffic-network and for forecasting mobility. However, these methods suffer from the inability to handle higher data dimensionality. The tensor-based approach has recently gained success over the existing methods due to its ability to decompose high dimension data into factor components. We present a modified Tucker decomposition method which predicts traffic mobility by approximating very large networks so as to handle the dimensionality problem. Our experiments on two big-city traffic-networks show that our method reduces the forecasting error, for up to 7 days, by around 80% as compared to the existing state of the art methods. Further, our method also efficiently handles the data dimensionality problem as compared to the existing methods. © 2018, Springer Nature Switzerland AG.

Abstract
Analytic approaches to combine interpretable models, although presented in different contexts, can be generalized to highlight the components that can be specialized. We propose a framework that structures the combination process, formalizes the problems that can be solved in alternative ways and evaluates the combined models based on their predictive ability to replace the base ones, without loss of interpretability. The framework is illustrated with a case study using data from the University of Porto, Portugal, where experiments were carried out. The results show that grouping base models by scientific areas, ordering by the number of variables and intersecting their underlying rules creates conditions for the combined models to outperform them. © 2019, Springer Nature Switzerland AG.