Steven Lemm

Publications of Steven Lemm

• Directional Variance Adjustment: improving covariance estimates for high-dimensional portfolio optimization

  Authors: Daniel Bartz, Kerr Hatrick, Christian W. Hesse, Klaus-Robert Müller, Steven Lemm

  09/2011;

  Robust and reliable covariance estimates play a decisive role in financial and many other applications. An important class of estimators is based on Factor models. Here, we show by extensive MonteRobust and reliable covariance estimates play a decisive role in financial and many other applications. An important class of estimators is based on Factor models. Here, we show by extensive Monte Carlo simulations that covariance matrices derived from the statistical Factor Analysis model exhibit a systematic error, which is similar to the well-known systematic error of the spectrum of the sample covariance matrix. Moreover, we introduce the Directional Variance Adjustment (DVA) algorithm, which diminishes the systematic error. In a thorough empirical study for the US, European, and Hong Kong market we show that our proposed method leads to improved portfolio allocation.

• Single-trial analysis and classification of ERP components--a tutorial.

  Authors: Benjamin Blankertz, Steven Lemm, Matthias Treder, Stefan Haufe, Klaus-Robert Müller

  NeuroImage. 05/2011; 56(2):814-25.

  Analyzing brain states that correspond to event related potentials (ERPs) on a single trial basis is a hard problem due to the high trial-to-trial variability and the unfavorable ratio between signalAnalyzing brain states that correspond to event related potentials (ERPs) on a single trial basis is a hard problem due to the high trial-to-trial variability and the unfavorable ratio between signal (ERP) and noise (artifacts and neural background activity). In this tutorial, we provide a comprehensive framework for decoding ERPs, elaborating on linear concepts, namely spatio-temporal patterns and filters as well as linear ERP classification. However, the bottleneck of these techniques is that they require an accurate covariance matrix estimation in high dimensional sensor spaces which is a highly intricate problem. As a remedy, we propose to use shrinkage estimators and show that appropriate regularization of linear discriminant analysis (LDA) by shrinkage yields excellent results for single-trial ERP classification that are far superior to classical LDA classification. Furthermore, we give practical hints on the interpretation of what classifiers learned from the data and demonstrate in particular that the trade-off between goodness-of-fit and model complexity in regularized LDA relates to a morphing between a difference pattern of ERPs and a spatial filter which cancels non task-related brain activity.

• Introduction to machine learning for brain imaging.

  Authors: Steven Lemm, Benjamin Blankertz, Thorsten Dickhaus, Klaus-Robert Müller

  NeuroImage. 05/2011; 56(2):387-99.

  Machine learning and pattern recognition algorithms have in the past years developed to become a working horse in brain imaging and the computational neurosciences, as they are instrumental forMachine learning and pattern recognition algorithms have in the past years developed to become a working horse in brain imaging and the computational neurosciences, as they are instrumental for mining vast amounts of neural data of ever increasing measurement precision and detecting minuscule signals from an overwhelming noise floor. They provide the means to decode and characterize task relevant brain states and to distinguish them from non-informative brain signals. While undoubtedly this machinery has helped to gain novel biological insights, it also holds the danger of potential unintentional abuse. Ideally machine learning techniques should be usable for any non-expert, however, unfortunately they are typically not. Overfitting and other pitfalls may occur and lead to spurious and nonsensical interpretation. The goal of this review is therefore to provide an accessible and clear introduction to the strengths and also the inherent dangers of machine learning usage in the neurosciences.

• Single-trial analysis and classification of ERP components -- a tutorial

  Authors: Benjamin Blankertz, Steven Lemm, Matthias Sebastian Treder, Stefan Haufe, Klaus-Robert Müller

  neuroimage. 01/2010;

• A generalized framework for quantifying the dynamics of EEG event-related desynchronization.

  Authors: Steven Lemm, Klaus-Robert Müller, Gabriel Curio

  PLoS computational biology. 09/2009; 5(8):e1000453.

  Brains were built by evolution to react swiftly to environmental challenges. Thus, sensory stimuli must be processed ad hoc, i.e., independent--to a large extent--from the momentary brain stateBrains were built by evolution to react swiftly to environmental challenges. Thus, sensory stimuli must be processed ad hoc, i.e., independent--to a large extent--from the momentary brain state incidentally prevailing during stimulus occurrence. Accordingly, computational neuroscience strives to model the robust processing of stimuli in the presence of dynamical cortical states. A pivotal feature of ongoing brain activity is the regional predominance of EEG eigenrhythms, such as the occipital alpha or the pericentral mu rhythm, both peaking spectrally at 10 Hz. Here, we establish a novel generalized concept to measure event-related desynchronization (ERD), which allows one to model neural oscillatory dynamics also in the presence of dynamical cortical states. Specifically, we demonstrate that a somatosensory stimulus causes a stereotypic sequence of first an ERD and then an ensuing amplitude overshoot (event-related synchronization), which at a dynamical cortical state becomes evident only if the natural relaxation dynamics of unperturbed EEG rhythms is utilized as reference dynamics. Moreover, this computational approach also encompasses the more general notion of a "conditional ERD," through which candidate explanatory variables can be scrutinized with regard to their possible impact on a particular oscillatory dynamics under study. Thus, the generalized ERD represents a powerful novel analysis tool for extending our understanding of inter-trial variability of evoked responses and therefore the robust processing of environmental stimuli.

• A novel mechanism for evoked responses in the human brain.

  Authors: Vadim V Nikulin, Klaus Linkenkaer-Hansen, Guido Nolte, Steven Lemm, Klaus R Müller, Risto J Ilmoniemi, Gabriel Curio

  The European journal of neuroscience. 06/2007; 25(10):3146-54.

  Magnetoencephalographic and electroencephalographic evoked responses are primary real-time objective measures of cognitive and perceptual processes in the human brain. Two mechanisms (additiveMagnetoencephalographic and electroencephalographic evoked responses are primary real-time objective measures of cognitive and perceptual processes in the human brain. Two mechanisms (additive activity and phase reset) have been debated and considered as the only possible explanations for evoked responses. Here we present theoretical and empirical evidence of a third mechanism contributing to the generation of evoked responses. Interestingly, this mechanism can be deduced entirely from the characteristics of spontaneous oscillations in the absence of stimuli. We show that the amplitude fluctuations of neuronal alpha oscillations at rest are associated with changes in the mean value of ongoing activity in magnetoencephalography, a phenomenon that we term baseline shifts associated with alpha oscillations. When stimuli modulate the amplitude of alpha oscillations, baseline shifts become the basis of a novel mechanism for the generation of evoked responses; the averaging of several trials leads to a cancellation of the oscillatory component but the baseline shift remains, which gives rise to an evoked response. We propose that the presence of baseline shifts associated with alpha oscillations can be explained by the asymmetric flow of inward and outward neuronal currents related to the generation of alpha oscillations. Our findings are relevant to the vast majority of electroencephalographic and magnetoencephalographic studies involving perceptual, cognitive and motor activity.

• Enhancing the signal-to-noise ratio of ICA-based extracted ERPs.

  Authors: Steven Lemm, Gabriel Curio, Yevhen Hlushchuk, Klaus-Robert Müller

  IEEE transactions on bio-medical engineering. 05/2006; 53(4):601-7.

  When decomposing single trial electroencephalography it is a challenge to incorporate prior physiological knowledge. Here, we develop a method that uses prior information about the phase-lockingWhen decomposing single trial electroencephalography it is a challenge to incorporate prior physiological knowledge. Here, we develop a method that uses prior information about the phase-locking property of event-related potentials in a regularization framework to bias a blind source separation algorithm toward an improved separation of single-trial phase-locked responses in terms of an increased signal-to-noise ratio. In particular, we suggest a transformation of the data, using weighted average of the single trial and trial-averaged response, that redirects the focus of source separation methods onto the subspace of event-related potentials. The practical benefit with respect to an improved separation of such components from ongoing background activity and extraneous noise is first illustrated on artificial data and finally verified in a real-world application of extracting single-trial somatosensory evoked potentials from multichannel EEG-recordings.

• Now you feel it--now you don't: ERP correlates of somatosensory awareness.

  Authors: Ruth Schubert, Felix Blankenburg, Steven Lemm, Arno Villringer, Gabriel Curio

  Psychophysiology. 02/2006; 43(1):31-40.

  We investigated correlates of somatosensory awareness for supra theshold stimuli using event-related potentials in a masking paradigm: Conscious perception of a weak, but suprathreshold "target"We investigated correlates of somatosensory awareness for supra theshold stimuli using event-related potentials in a masking paradigm: Conscious perception of a weak, but suprathreshold "target" stimulus was suppressed in a significant number of trials when followed by a higher-intensity "mask" stimulus. ERPs were compared for trials with perceived versus unperceived target stimuli. Early ERPs (P60, N80), generated in the contralateral S1, were found independent of stimulus perception. In contrast, for consciously perceived target stimuli, amplitude enhancements were observed for the P100 and N140. Thus, early activation of S1 is not sufficient to warrant conscious stimulus perception. Conscious stimulus processing differs significantly from unconscious processing starting around 100 ms after stimulus presentation when the signal is processed in parietal and frontal cortices, brain regions crucial for stimulus access into conscious perception.

• Aggregating Classification Accuracy across Time: Application to Single Trial EEG.

  Authors: Steven Lemm, Christin Schäfer, Gabriel Curio

  Advances in Neural Information Processing Systems 19, Proceedings of the Twentieth Annual Conference on Neural Information Processing Systems, Vancouver, British Columbia, Canada, December 4-7, 2006; 01/2006

• The Berlin Brain-Computer Interface: Machine Learning Based Detection of User Specific Brain States.

  Authors: Benjamin Blankertz, Guido Dornhege, Steven Lemm, Matthias Krauledat, Gabriel Curio, Klaus-Robert Müller

  J. UCS. 01/2006; 12:581-607.

• Binary on-line classification based on temporally integrated information

  Authors: Christin Schäfer, Steven Lemm, Gabriel Curio

  01/2005

• BCI Competition 2003--Data set III: probabilistic modeling of sensorimotor mu rhythms for classification of imaginary hand movements.

  Authors: Steven Lemm, Christin Schäfer, Gabriel Curio

  IEEE transactions on bio-medical engineering. 07/2004; 51(6):1077-80.

  Brain-computer interfaces require effective online processing of electroencephalogram (EEG) measurements, e.g., as a part of feedback systems. We present an algorithm for single-trial onlineBrain-computer interfaces require effective online processing of electroencephalogram (EEG) measurements, e.g., as a part of feedback systems. We present an algorithm for single-trial online classification of imaginary left and right hand movements, based on time-frequency information derived from filtering EEG wideband raw data with causal Morlet wavelets, which are adapted to individual EEG spectra. Since imaginary hand movements lead to perturbations of the ongoing pericentral mu rhythm, we estimate probabilistic models for amplitude modulation in lower (10 Hz) and upper (20 Hz) frequency bands over the sensorimotor hand cortices both contra- and ipsilaterally to the imagined movements (i.e., at EEG channels C3 and C4). We use an integrative approach to accumulate over time evidence for the subject's unknown motor intention. Disclosure of test data labels after the competition showed this approach to succeed with an error rate as low as 10.7%.

• Enhancing Brain-Computer Interfaces by Machine Learning Techniques Benjamin Blankertz

  Authors: Benjamin Blankertz, Guido Dornhege, Steven Lemm, Gabriel Curio

  02/2004;

  movements, the monkey can control a robotic arm by brain activity alone. Figure taken from [Nicolelis et al, 2000] [Laubach et al., 2000, Kennedy et al., 2000, Reina et al., 2001, Levine et al.,movements, the monkey can control a robotic arm by brain activity alone. Figure taken from [Nicolelis et al, 2000] [Laubach et al., 2000, Kennedy et al., 2000, Reina et al., 2001, Levine et al., 2000] BCI using SSVEP Regarding a stimulus blinking at a frequency between 7 and 30 Hz evokes a rhythm of the same frequency in the visual cortex. 2 3 4 6 7 8 5 1 9 0 In the Beijing setup each button flashes at an individual frequency. By spectral analysis of the EEG the regarded button can be detected from 1 s windows. [dependent, asynchronous, evoked potentials [Middendorf et al., 2000, Cheng et al., 2002] BCI using the P300 Component An awaited infrequent stimulus (deviant) in a series of standard stimuli elicits a P300 component at central scalp position. In the Donchin setup the subject concentrates on a letter of a 66 symbol matrix. Rows and columns are highlighted several times in random order. P300 components are most strongly elicited when the row resp. column is flashed which con

• Spatio-Spectral Filters for Improving the Classi cation of Single Trial EEG

  Authors: Steven Lemm, Benjamin Blankertz, Gabriel Curio, Klaus-robert Müller

  October. 52(9):1541-1548.

• Spatio-spectral filters for improving the classification of single trial EEG.

  Authors: Steven Lemm, Benjamin Blankertz, Gabriel Curio, Klaus-Robert Müller

  IEEE transactions on bio-medical engineering. 52(9):1541-8.

  Data recorded in electroencephalogram (EEG)-based brain-computer interface experiments is generally very noisy, non-stationary, and contaminated with artifacts that can deteriorateData recorded in electroencephalogram (EEG)-based brain-computer interface experiments is generally very noisy, non-stationary, and contaminated with artifacts that can deteriorate discrimination/classification methods. In this paper, we extend the common spatial pattern (CSP) algorithm with the aim to alleviate these adverse effects. In particular, we suggest an extension of CSP to the state space, which utilizes the method of time delay embedding. As we will show, this allows for individually tuned frequency filters at each electrode position and, thus, yields an improved and more robust machine learning procedure. The advantages of the proposed method over the original CSP method are verified in terms of an improved information transfer rate (bits per trial) on a set of EEG-recordings from experiments of imagined limb movements.