Smart devices, such as smartphones, wearables, robots, and others, can collect vast amounts of data from their environment. This data is suitable for training machine learning models, which can significantly improve their behavior, and therefore, the user experience. Federated learning is a young and popular framework that allows multiple distributed devices to train deep learning models collaboratively while preserving data privacy. Nevertheless, this approach may not be optimal for scenarios where data distribution is non-identical among the participants or changes over time, causing what is known asconcept drift. Little research has yet been done in this field, but this kind of situation is quite frequent in real life and poses new challenges to both continual and federated learning. Therefore, in this work, we present a new method, called Concept-Drift-Aware Federated Averaging (CDA-FedAvg). Our proposal is an extension of the most popular federated algorithm, Federated Averaging (FedAvg), enhancing it for continual adaptation under concept drift. We empirically demonstrate the weaknesses of regular FedAvg and prove that CDA-FedAvg outperforms it in this type of scenario.

N this article we describe the applica- I tion of a new artificial neural network model aimed at the morphological classi- fication of heartbeats detected on a multi- channel ECG signal. We emphasize the special characteristics of the algorithm as an adaptive classifier with the capacity to dynamically self-organize its response to the characteristics of the ECG input sig- nal. We also present evaluation results based on traces from the MIT-BIH ar- rhythmia database. Importance of Classifying the QRS Complex Analysis of the ECG signal is of great importance in the detection of cardiac anomalies. One of the most important signal intervals in this respect is the QRS complex, which is associated with elec- trical ventricular activation. Besides other information contained in this signal segment, its principal application is to be found in the important modifications that its morphology undergoes in beats of ventricular origin. Consequently, the morphological classification of QRS complexes (henceforth QRSs) consti- tutes an important step in the detection of ventricular arrhythmias, the solution of which, apart from constituting another pattern-recognition problem, has certain characteristics that should be described. First, the number and variety of the possi- ble morphological classes is practically infinite. The appearance of heartbeats varies considerably, not only among pa- tients, but also as the consequences of movement, modifications in the electri- cal characteristics of the body, etc. Thus, it is not possible to construct a training set that is representative of all the possi- ble morphologies. On the other hand, the ECG signal fre- quently presents noise components of various origins; among these one could point out the biological ones, such as, for example, the movement of patients' mus- cular masses, which generates high-fre-

Federated learning is the state-of-the-art paradigm for training a learning model collaboratively across multiple distributed devices while ensuring data privacy. Under this framework, different algorithms have been developed in recent years and have been successfully applied to real use cases. The vast majority of work in federated learning assumes static datasets and relies on the use of deep neural networks. However, in real-world problems, it is common to have a continual data stream, which may be non-stationary, leading to phenomena such as concept drift. Besides, there are many multi-device applications where other, non-deep strategies are more suitable, due to their simplicity, explainability, or generalizability, among other reasons. In this paper we present Ensemble and Continual Federated Learning, a federated architecture based on ensemble techniques for solving continual classification tasks. We propose the global federated model to be an ensemble, consisting of several independent learners, which are locally trained. Thus, we enable a flexible aggregation of heterogeneous client models, which may differ in size, structure, or even algorithmic family. This ensemble-based approach, together with drift detection and adaptation mechanisms, also allows for continual adaptation in situations where data distribution changes over time. In order to test our proposal and illustrate how it works, we have evaluated it in different tasks related to human activity recognition using smartphones.