@article {18547, title = {Anomaly and Change Detection in Graph Streams through Constant-Curvature Manifold Embeddings}, journal = {IJCNN 2018 : International Joint Conference on Neural Networks}, year = {2018}, month = {07/2018}, author = {Zambon, Daniele and Livi, Lorenzo and Alippi, Cesare} } @conference {18593, title = {A characterization of the Edge of Criticality in Binary Echo State Networks}, booktitle = {2018 IEEE 28th International Workshop on Machine Learning for Signal Processing (MLSP)}, year = {2018}, month = {09/2018}, address = {Aalborg, Denmark}, author = {Verzelli, Pietro and Livi, Lorenzo and Alippi, Cesare} } @article {18545, title = {Concept Drift and Anomaly Detection in Graph Streams}, journal = {IEEE Transactions on Neural Networks and Learning Systems}, volume = {29}, issue = {11}, year = {2018}, month = {11/2018}, pages = {5592-5605}, chapter = {5592}, author = {Zambon, Daniele and Alippi, Cesare and Livi, Lorenzo} } @article {18533, title = {Determination of the Edge of Criticality in Echo State Networks Through Fisher Information Maximization}, journal = {IEEE Transactions on Neural Networks and Learning Systems}, volume = {29}, year = {2018}, month = {March}, pages = {706-717}, keywords = {Echo state network (ESN), echo state networks, edge of criticality, ESN, Fisher information, Fisher information matrix, Fisher information maximization, Jacobian matrices, Learning systems, Neurons, nonparametric estimation, nonparametric estimator, nonparametric statistics, prediction error, Probability density function, recurrent neural nets, recurrent neural networks, Reservoirs, RNN, short-term memory capacity, Training}, issn = {2162-237X}, doi = {10.1109/TNNLS.2016.2644268}, author = {Livi, Lorenzo and Bianchi, Filippo Maria and Alippi, Cesare} } @article {18508, title = {Investigating echo state networks dynamics by means of recurrence analysis}, journal = {IEEE Transactions on Neural Networks and Learning Systems}, volume = {29}, year = {2018}, month = {02/2018}, pages = { 427 - 439}, abstract = {In this paper, we elaborate over the well-known interpretability issue in echo state networks. The idea is to investigate the dynamics of reservoir neurons with time-series analysis techniques taken from research on complex systems. Notably, we analyze time-series of neuron activations with Recurrence Plots (RPs) and Recurrence Quantification Analysis (RQA), which permit to visualize and characterize high-dimensional dynamical systems. We show that this approach is useful in a number of ways. First, the two-dimensional representation offered by RPs provides a way for visualizing the high-dimensional dynamics of a reservoir. Our results suggest that, if the network is stable, reservoir and input denote similar line patterns in the respective RPs. Conversely, the more unstable the ESN, the more the RP of the reservoir presents instability patterns. As a second result, we show that the Lmax measure is highly correlated with the well-established maximal local Lyapunov exponent. This suggests that complexity measures based on RP diagonal lines distribution provide a valuable tool to quantify the degree of network stability. Finally, our analysis shows that all RQA measures fluctuate on the proximity of the so-called edge of stability, where an ESN typically achieves maximum computational capability. We verify that the determination of the edge of stability provided by such RQA measures is more accurate than two well-known criteria based on the Jacobian matrix of the reservoir. Therefore, we claim that RPs and RQA-based analyses can be used as valuable tools to design an effective network given a specific problem.}, doi = {10.1109/TNNLS.2016.2630802}, author = {Bianchi, Filippo Maria and Livi, Lorenzo and Alippi, Cesare} } @conference {18530, title = {Time series kernel similarities for predicting Paroxysmal Atrial Fibrillation from ECGs}, booktitle = { IJCNN 2018 : International Joint Conference on Neural Networks}, year = {2018}, month = {07/2018}, publisher = {IEEE}, organization = {IEEE}, address = {Rio, Brazil}, author = {Bianchi, Filippo Maria and Livi, Lorenzo and Ferrante, Alberto and Milosevic, Jelena and Malek, Miroslaw} } @conference {18538, title = {Critical echo state network dynamics by means of Fisher information maximization}, booktitle = {2017 International Joint Conference on Neural Networks (IJCNN)}, year = {2017}, month = {May}, keywords = {Asymptotic stability, critical phase transitions, echo state network, edge of criticality, Electronic mail, ESN, Estimation, estimation theory, FIM, Fisher information matrix, Fisher information maximization, hidden neurons activations, high short term memory capacity, input signal, low prediction error, network dynamics, network theory (graphs), Neurons, optimisation, Probability density function, recurrent neural nets, Reservoirs, signal processing, statistics, Tuning, unsupervised approach, unsupervised learning}, doi = {10.1109/IJCNN.2017.7965941}, author = {Bianchi, Filippo Maria and Livi, Lorenzo and Jenssen, Robert and Alippi, Cesare} } @conference {18542, title = {Detecting changes in sequences of attributed graphs}, booktitle = {2017 IEEE Symposium Series on Computational Intelligence (SSCI)}, year = {2017}, month = {Nov}, keywords = {Aerospace electronics, Anomaly detection, application domains, Attributed graph, attributed graphs, Change detection, Concept drift, Dynamic/Evolving graph, Electronic mail, Embedding, geometric graphs, Graph matching, graph theory, graph-based representations, Markov chains, Markov processes, Microsoft Windows, pair-wise relations, Prototypes, real-world systems, Stationarity, topology, Training, variable order}, doi = {10.1109/SSCI.2017.8285273}, author = {Zambon, Daniele and Livi, Lorenzo and Alippi, Cesare} } @article {18532, title = {Multiplex visibility graphs to investigate recurrent neural network dynamics}, journal = {Nature-Scientific reports}, volume = {7}, year = {2017}, month = {03/2017}, pages = {44037-44049}, chapter = {44037}, doi = {doi:10.1038/srep44037}, author = {Bianchi, Filippo Maria and Livi, Lorenzo and Alippi, Cesare and Jenssen, Robert} } @article {18510, title = {One-class classifiers based on entropic spanning graphs}, journal = {IEEE Transactions on Neural Networks and Learning Systems}, volume = {28}, issue = {12}, year = {2017}, month = {11/2016}, pages = { 2846 - 2858}, chapter = {2846}, abstract = {One-class classifiers offer valuable tools to assess the presence of outliers in data. In this paper, we propose a design methodology for one-class classifiers based on entropic spanning graphs. The spanning graph is learned on the embedded input data, with the aim to generate a partition of the vertices. The final partition is derived by exploiting a criterion based on mutual information minimization. Here, we compute the mutual information by using a convenient formulation provided in terms of the alfa-Jensen difference. Once training is completed, in order to associate a confidence level with the classifier decision, a graph-based fuzzy model is constructed. The fuzzification process is based only on topological information of the vertices of the entropic spanning graph. As such, the proposed one-class classifier is suitable also for datasets with complex geometric structures. We provide experiments on well-known benchmarking datasets containing both feature vectors and labeled graphs. In addition, we apply the method on the problem of protein solubility recognition by considering several data representations for the samples. Experimental results demonstrate the effectiveness and versatility of the proposed method with respect to other state-of the-art approaches. }, doi = { 10.1109/TNNLS.2016.2608983}, author = {Livi, Lorenzo and Alippi, Cesare} } @conference {18448, title = {One-Class Classification Through Mutual Information Minimization}, booktitle = {IEEE-INNS International Joint Conference on Neural Networks (IJCNN16)}, year = {2016}, month = {07/2016}, address = {Vancouver, Canada}, abstract = {In one-class classification problems, a model is synthesized by using only information coming from the nominal state of the data generating process. Many important applications can be cast in the one-class classification framework, such as anomaly and change in stationarity detection, and fault recognition. In this paper, we present a novel design methodology for oneclass classifiers derived from graph-based entropy estimators. The entropic graph is used to generate a partition of the input nominal conditions, which corresponds to the classifier model. Here we propose a criterion based on mutual information minimization to learn such a partition. The _-Jensen difference is considered, which provides a convenient way for estimating the mutual information. The classifier incorporates also a fuzzy model, providing a confidence value for a generic test sample during operational modality, expressed as a membership degree of the sample to the nominal conditions class. The fuzzification mechanism is based only on topological properties of the entropic spanning graph vertices; as such, it allows to model clusters of arbitrary shapes. We show preliminary {\textendash} yet very promising {\textendash} results on both synthetic problems and real-world datasets for one-class classification. }, author = {Livi, Lorenzo and Alippi, Cesare} }