Large data is challenging for most existing discovery algorithms, for several reasons. First of all, such data leads to enormous hypothesis spaces, making exhaustive search infeasible. Second, many variants of essenti...
详细信息
We consider the problem of estimating brain effective connectivity from electroencephalographic (EEG) measurements, which is challenging due to instantaneous correlations in the sensor data caused by volume conduction...
详细信息
In this paper, a family of support vector novelty detection (or SVND) in hidden space is presented. Firstly a hidden-space SVND (or HSVND) algorithm is proposed. The data in an input space is mapped into a hidden spac...
详细信息
Most present-day visual brain computer interfaces (BCIs) suffer from the fact that they rely on eye movements, are slow-paced, or feature a small vocabulary. As a potential remedy, we explored a novel BCI paradigm con...
详细信息
Most present-day visual brain computer interfaces (BCIs) suffer from the fact that they rely on eye movements, are slow-paced, or feature a small vocabulary. As a potential remedy, we explored a novel BCI paradigm consisting of a central rapid serial visual presentation (RSVP) of the stimuli. It has a large vocabulary and realizes a BCI system based on covert non-spatial selective visual attention. In an offline study, eight participants were presented sequences of rapid bursts of symbols. Two different speeds and two different color conditions were investigated. Robust early visual and P300 components were elicited time-locked to the presentation of the target. Offline classification revealed a mean accuracy of up to 90% for selecting the correct symbol out of 30 possibilities. The results suggest that RSVP-BCI is a promising new paradigm, also for patients with oculomotor impairments.
In this paper, we investigate the use of event-related potentials (ERPs) as a quantitative measure for quality assessment of disturbed audio signals. For this purpose, we ran an EEG study (N=11) using an oddball parad...
详细信息
In this paper, we investigate the use of event-related potentials (ERPs) as a quantitative measure for quality assessment of disturbed audio signals. For this purpose, we ran an EEG study (N=11) using an oddball paradigm, during which subjects were presented with the phoneme /a/, superimposed with varying degrees of signal-correlated noise. Based on this data set, we address the question to which degree the degradation of the auditory stimuli is reflected on a neural level, even if the disturbance is below the threshold of conscious perception. For those stimuli that are consciously recognized as being disturbed, we suggest the use of the amplitude and latency of the P300 component for assessing the level of disturbance. For disturbed stimuli for which the noise is not perceived consciously, we show for two subjects that a classifier based on shrinkage LDA can be applied successfully to single out stimuli, for which the noise was presumably processed subconsciously.
EEG single-trial analysis requires methods that are robust against noise and disturbance. In this contribution, based on the framework of robust statistics, we propose a simple modification of Common Spatial Patterns ...
详细信息
Recently, a novel subspace decomposition method, termed 'Stationary Subspace analysis' (SSA), has been proposed by Bünau et al. [10]. SSA aims to find a linear projection to a lower dimensional subspace s...
详细信息
Recently, a novel subspace decomposition method, termed `Stationary Subspace analysis' (SSA), has been proposed by Bu¿nau et al.. SSA aims to find a linear projection to a lower dimensional subspace such that...
详细信息
Recently, a novel subspace decomposition method, termed `Stationary Subspace analysis' (SSA), has been proposed by Bu¿nau et al.. SSA aims to find a linear projection to a lower dimensional subspace such that the distribution of the projected data does not change over successive epochs or sub-datasets. We show that by modifying the loss function and the optimization procedure we can obtain an algorithm that is both faster and more accurate. We discuss the problem of indeterminacies and provide a lower bound on the number of epochs that is needed. Finally, we show in an experiment with simulated image patches, that SSA can be used favourably in invariance learning.
暂无评论