Search

Objective: Recent studies have used sparse classifications to predict categorical variables from high-dimensional brain activity signals to expose human's mental states and intentions, selecting the relevant features automatically in the model training process. However, existing sparse classification models will likely be prone to the performance degradation which is caused by the noise inherent in the brain recordings. To address this issue, we aim to propose a new robust and sparse classification algorithm in this study. Methods: To this end, we introduce the correntropy learning framework into the automatic relevance determination based sparse classification model, proposing a new correntropy-based robust sparse logistic regression algorithm. To demonstrate the superior brain activity decoding performance of the proposed algorithm, we evaluate it on a synthetic dataset, an electroencephalogram (EEG) dataset, and a functional magnetic resonance imaging (fMRI) dataset. Results: The extensive experimental results confirm that not only the proposed method can achieve higher classification accuracy in a noisy and high-dimensional classification task, but also it would select those more informative features for the decoding tasks. Conclusion: Integrating the correntropy learning approach with the automatic relevance determination technique will significantly improve the robustness with respect to the noise, leading to more adequate robust sparse brain decoding algorithm. Significance: It provides a more powerful approach in the real-world brain activity decoding and the brain-computer interfaces.

The minimum error entropy (MEE) criterion is a powerful approach for non-Gaussian signal processing and robust machine learning. However, the instantiation of MEE on robust classification is a rather vacancy in the literature. The original MEE purely focuses on minimizing Renyi’s quadratic entropy of the prediction errors, which could exhibit inferior capability in noisy classification tasks. To this end, we analyze the optimal error distribution with adverse outliers and introduce a specific codebook for restriction, which optimizes the error distribution toward the optimal case. Half-quadratic-based optimization and convergence analysis of the proposed learning criterion, called restricted MEE (RMEE), are provided. The experimental results considering logistic regression and extreme learning machine on synthetic data and UCI datasets, respectively, are presented to demonstrate the superior robustness of RMEE. Furthermore, we evaluate RMEE on a noisy electroencephalogram dataset, so as to strengthen its practical impact.

This study proposes a van der Pol-like complementary chaotic oscillator circuit. By exploiting components from both semiconductor and vacuum tube technologies, specifically field-effect transistors and triodes, different physical realizations of this oscillator are achieved. The proposed design is an asymmetric cross-coupled complementary circuit derived from the classical van der Pol oscillator coupled to a linear circuit. This study explores the unique physical properties of the nonlinear amplifying elements and the resulting dynamic behaviors of the systems, systematically comparing their characteristics from a dynamical perspective and initially investigating their applicability to physiological data augmentation. First, a detailed analysis of the effects of the nonlinearities of each device on the system behavior is performed, revealing that triodes lead to smoother and more stable oscillations, including weak chaos due to torus breakdown, whereas field effect transistors lead to more disordered chaotic oscillations, accompanied by strong hysteresis, coexistence of attractors, and unstable oscillations. Then, for each amplifier device, the corresponding oscillator is driven by representative physiological signals, including electroencephalograms, electrocardiograms, electromyograms, and plethysmograms, thus harnessing the system response to the disturbance and gaining insights into its ability to enhance data, qualitatively through spectrograms and quantitatively by means of information theoretical measurements. This work illustrates the possibility of harnessing the nonidealities of diverse existing physical devices towards nonlinear signal processing. • A van der Pol-like complementary chaotic oscillator was designed and realized using triodes and JFETs transistors. • The qualitative features of the dynamics were largely consistent across the equation and SPICE models and experiments. • Triodes produced stable, smooth oscillation with weak chaos, whereas FETs more disordered activity. • The side-by-side comparison of different circuit realizations illustrated the effect of physical non-idealities. • The possible relevance of circuits of this kind in the data augmentation of physiological signals (EEG, ECG) was exemplified. [ABSTRACT FROM AUTHOR]

Coupled non-linear dynamical systems have gained attention not only as a ubiquitous occurrence in natural and artificial scenarios, but also as a basis for atypical computation paradigms. This paper introduces an approach to time series data augmentation involving driving a single low-dimensional entity, namely the Rössler system, with a physically-recorded sensor signal, and leveraging its responses to enhance the performance of a conventional classifier. A representative internet of things application in agriculture, namely cattle behavior recognition using a triaxial accelerometer, is investigated via a publicly-available dataset. Numerical simulations and experiments with an analog electronic circuit reveal that diversified responses to the external input are attainable, and the additional time series obtained from the driven system enhance the behavior classification accuracy. The advantage, down to the combined effects of its dynamical response and a static non-linearity transforming the driving signal, is appreciable both when using a small multi-layer perceptron network operating on elementary features and, albeit to a lesser extent, when feeding the time series directly to a convolutional neural network. One possibility is that the driven system translates non-linear dynamical features into linear signal properties that can be more easily extracted. Some considerations about the engineering implementation using either analog hardware or programmable logic on an edge device are given. • Accelerometer recordings are entered via a forcing term into simulated Rössler systems. • The responses of the Rössler systems are additionally supplied to classifier networks. • A relevant increase in animal behavior classification accuracy is therefore attained. • The resulting form of data augmentation is also demonstrated using an analog circuit. • Low-power implementation suitable for deployment on edge devices appears feasible. [ABSTRACT FROM AUTHOR]