This paper proposes an algorithm called Forward Composition Propagation (FCP) to explain the predictions of feed-forward neural networks operating on structured pattern recognition problems. In the proposed FCP algorithm, each neuron is described by a composition vector indicating the role of each problem feature in that neuron. Composition vectors are initialized using a given input instance and subsequently propagated through the whole network until we reach the output layer. It is worth mentioning that the algorithm is executed once the network's training network is done. The sign of each composition value indicates whether the corresponding feature excites or inhibits the neuron, while the absolute value quantifies such an impact. Aiming to validate the FCP algorithm's correctness, we develop a case study concerning bias detection in a state-of-the-art problem in which the ground truth is known. The simulation results show that the composition values closely align with the expected behavior of protected features.

Forecasting windmill time series is often the basis of other processes such as anomaly detection, health monitoring, or maintenance scheduling. The amount of data generated on windmill farms makes online learning the most viable strategy to follow. Such settings require retraining the model each time a new batch of data is available. However, update the model with the new information is often very expensive to perform using traditional Recurrent Neural Networks (RNNs). In this paper, we use Long Short-term Cognitive Networks (LSTCNs) to forecast windmill time series in online settings. These recently introduced neural systems consist of chained Short-term Cognitive Network blocks, each processing a temporal data chunk. The learning algorithm of these blocks is based on a very fast, deterministic learning rule that makes LSTCNs suitable for online learning tasks. The numerical simulations using a case study with four windmills showed that our approach reported the lowest forecasting errors with respect to a simple RNN, a Long Short-term Memory, a Gated Recurrent Unit, and a Hidden Markov Model. What is perhaps more important is that the LSTCN approach is significantly faster than these state-of-the-art models.

The Millennium Institute for Foundational Research on Dataa (IMFD) started its operations in June 2018, funded by the Millennium Science Initiative of the Chilean National Agency of Research and Development.b IMFD is a joint initiative led by Universidad de Chile and Universidad Católica de Chile, with the participation of five other Chilean universities: Universidad de Concepción, Universidad de Talca, Universidad Técnica Federico Santa María, Universidad Diego Portales, and Universidad Adolfo Ibáñez. IMFD aims to be a reference center in Latin America related to state-of-the-art research on the foundational problems with data, as well as its applications to tackling diverse issues ranging from scientific challenges to complex social problems. As tasks of this kind are interdisciplinary by nature, IMFD gathers a large number of researchers in several areas that include traditional computer science areas such as data management, Web science, algorithms and data structures, privacy and verification, information retrieval, data mining, machine learning, and knowledge representation, as well as some areas from other fields, including statistics, political science, and communication studies. IMFD currently hosts 36 researchers, seven postdoctoral fellows, and more than 100 students.