The non-negative matrix factorization algorithm (NMF) decomposes a data matrix into a set of non-negative basis vectors, each scaled by a coefficient. In its original formulation, the NMF assumes the data samples and ...
详细信息
ISBN:
(纸本)9781424492695
The non-negative matrix factorization algorithm (NMF) decomposes a data matrix into a set of non-negative basis vectors, each scaled by a coefficient. In its original formulation, the NMF assumes the data samples and dimensions to be independently distributed, making it a less-than-ideal algorithm for the analysis of time series data with temporal correlations. Here, we seek to derive an NMF that accounts for temporal dependencies in the data by explicitly incorporating a very simple temporal constraint for the coefficients into the NMF update rules. We applied the modified algorithm to 2 multi-dimensional electromyographic data sets collected from the human upper-limb to identify muscle synergies. We found that because it reduced the number of free parameters in the model, our modified NMF made it possible to use the Akaike Information Criterion to objectively identify a model order (i.e., the number of muscle synergies composing the data) that is more functionally interpretable, and closer to the numbers previously determined using ad hoc measures.
We analyze minimization algorithms for L-q-convex functions in discrete convex analysis and establish exact bounds for the number of iterations required by the steepest descent algorithm and its variants. (C) 2014 Els...
详细信息
We analyze minimization algorithms for L-q-convex functions in discrete convex analysis and establish exact bounds for the number of iterations required by the steepest descent algorithm and its variants. (C) 2014 Elsevier B.V. All rights reserved.
Technology scaling is a common trend in current embedded systems. It has promoted the use of multi-core, multiprocessor, and distributed platforms. Such systems usually require run-time migration of distributed applic...
详细信息
ISBN:
(纸本)9781450320719
Technology scaling is a common trend in current embedded systems. It has promoted the use of multi-core, multiprocessor, and distributed platforms. Such systems usually require run-time migration of distributed applications between the different nodes of the platform in order to balance the workload or to tolerate faults. Before an application can be migrated, it needs to be brought to a stable state such that restarting the application after migration does not violate its functional correctness. An application in a stable state does not change its context any further, and therefore, stabilization is a prerequisite for any application migration. Process networks are a common model of computation for specifying distributed applications. However, most results on the migration of process networks do not provide an algorithm to put a general process network into a stable state, suitable for migration. This paper proposes a technique which efficiently and correctly brings a process network executing on a distributed system to a known stable state. The correctness of the technique is independent of the temporal characteristics of the system and the topology of the process network. The required modifications of a process network are lightweight and preserve its original functionality. A model characterizing the timing properties of the technique is provided. The feasibility and efficiency of the proposed approach and the respective model are validated with experimental results on Intel's SCC platform.
We validate a novel algorithm to detect saccades from raw data obtained during walking from a mobile infra-red eye-tracking device. The algorithm was based on a velocity threshold detection method, which excluded arte...
详细信息
ISBN:
(纸本)9781424479290
We validate a novel algorithm to detect saccades from raw data obtained during walking from a mobile infra-red eye-tracking device. The algorithm was based on a velocity threshold detection method, which excluded artefacts such as blinks and flickers using specific criteria. Mobile infra-red eye-tracking was performed with a group of healthy older adults (n=5) and Parkinson's disease (n=5) subjects. Saccades determined from raw eye tracker data obtained during walking using the algorithm were compared to a ground truth dataset defined as frame-by-frame visual inspection of raw eye-tracking videos. 100 trials from 10 subjects were analyzed and compared. The algorithm was highly reliable when compared to the ground truth (ICC(2,1) = 0.94), with an overall correct saccade detection percentage of 85%. This provides a simple yet robust algorithm for the analysis of mobile eye-tracking data.
Various independent component analysis (ICA) algorithms (MILCA, JADE, SIMPLISMA, RADICAL) are applied for simultaneous spectroscopic determination of two groups of transition metals: Co(II)-Fe(III)Cu(II)-Zn(II)-Ni(II)...
详细信息
Various independent component analysis (ICA) algorithms (MILCA, JADE, SIMPLISMA, RADICAL) are applied for simultaneous spectroscopic determination of two groups of transition metals: Co(II)-Fe(III)Cu(II)-Zn(II)-Ni(II) and Pt(IV)-Pd(II)-Ir(IV)-Rh(III)-Ru(III)) in complex mixtures. The analysis is based on the decomposition of spectra of multicomponent mixtures in the UV-VIS region based on the natural absorbance of metal salts, or, when a better sensitivity is desirable, based on the absorbance of their complexes with 4-(2-pyridylazo) resorcinol (PAR) and ethylenediaminetetraacetic acid (EDTA). Good quality spectral resolution of up to seven-component mixtures was achieved (correlation coefficients between resolved and experimental spectra are not less than 0.90). In general, the relative errors in the recovered concentrations are at levels of only several percent. While being superior to other ICA algorithms, MILCA is comparable or even outperforms other classical chemometric methods for quantitative analysis that were used for comparison purposes (Partial Least Squares (PLS), Principal Component Regression (PCR), Alternating Least Squares (ALS)). Simultaneous quantitative analysis is possible for mixtures containing up to five metals in the broad concentration ranges even when individual spectra show 99% overlap. A small excess of derivatization reagent (till threefold excess to the sum of metal concentrations) is optimal to obtain good quantitative results. The proposed method was used for analysis of authentic samples (multimineral supplements and platinum concentrates). The resolved ICA concentrations match well with the labelled amounts and the results of other chemometric methods (ALS, PLS). ICA decomposition considerably improves the application range of spectroscopy for metal quantification in mixtures.
Automatic transcription of polyphonic music has been an active research field for several years and is considered by many to be a key enabling technology in music signal processing. However, current transcription appr...
详细信息
ISBN:
(纸本)9781479928941
Automatic transcription of polyphonic music has been an active research field for several years and is considered by many to be a key enabling technology in music signal processing. However, current transcription approaches either focus on detecting pitched sounds (from pitched musical instruments) or on detecting unpitched sounds (from drum kits). In this paper, we propose a method that jointly transcribes pitched and unpitched sounds from polyphonic music recordings. The proposed model extends the probabilistic latent component analysisalgorithm and supports the detection of pitched sounds from multiple instruments as well as the detection of unpitched sounds from drum kit components, including bass drums, snare drums, cymbals, hi-hats, and toms. Our experiments based on polyphonic Western music containing both pitched and unpitched instruments led to very encouraging results in multi-pitch detection and drum transcription tasks.
For multicast security routing in optical network, through the adoption of maximum light path attack radius MaxLAR criterion, Combined with the advantages of the immune algorithm and artificial fish algorithm, this pa...
详细信息
ISBN:
(纸本)9781510802704
For multicast security routing in optical network, through the adoption of maximum light path attack radius MaxLAR criterion, Combined with the advantages of the immune algorithm and artificial fish algorithm, this paper proposes a new attack-aware routing algorithm based on immune artificial fish swarm optimization in optical networks. The algorithm can find out the smallest light path attack radius in light paths under the given condition, can enhance the capacity of optical network attack prevention, can reduce the light damage by the potential physical-layer attacks. Through algorithmanalysis, we can show that the algorithm is feasible, and have achieved good performance.
MIMO techniques can significantly improve the diversity gain and spectrum utilization. Physical-layer Network Coding (PNC) is also a new technique to increase the spectrum utilization of wireless networks. The novel V...
详细信息
ISBN:
(纸本)9781479914036
MIMO techniques can significantly improve the diversity gain and spectrum utilization. Physical-layer Network Coding (PNC) is also a new technique to increase the spectrum utilization of wireless networks. The novel VBLAST PNC scheme combined these two techniques to improve the performance further. In this paper, we analyze the Bit Error Rate (BER) performance of the new scheme based on QPSK modulation over Rayleigh fading channels. Although the adaptive VBLAST PNC in the new scheme has a better BER performance, the analysis of this algorithm is too difficult to derive the closed-form expressions. Hence, we just present the exact analytical BER performance of VBLAST PNC in a near closed-form. The BER analysis is then verified by the numerical simulation. Inheriting the basic idea from VBLAST PNC scheme and with the help of analytical BER performance, we may develop a better PNC decoding strategy in further works.
New packaging technologies are necessary to meet the demand for smaller and more reliable electronic devices. The so-called system-in-package (SiP) concept brings different technologies, material combinations and proc...
详细信息
ISBN:
(纸本)9781479922710
New packaging technologies are necessary to meet the demand for smaller and more reliable electronic devices. The so-called system-in-package (SiP) concept brings different technologies, material combinations and processes together in one package. To ensure the reliability of such a package reliable, fast and non-destructive failure analysis are needed. Transient thermal methods provide techniques for detection of different types of defects such as delaminations of interfaces, electrical short circuits and voids in die attach layer [1]-[3]. Especially for 3D integration such failures occur inside the system and therefore they are not visible for some failure analytical tools. Scanning acoustic microscopy (SAM) and X-Ray analysis are well-established techniques with some disadvantages. For example, one want to avoid that components are placed in a water bath as it is necessary for SAM investigations. Transient thermal methods were successfully applied to detect very weak thermal responses of joules heat sources in similar to 1 mK range and temperature changes [1] and periodic changing stress field of a crack tip [3]. This work gives an overview of the different analysisalgorithms used in transient thermal investigations.
An efficient algorithm for analysis and optimization of rotationally symmetric reflector antennas, possibly including 3D support structures and/or waveguide components, is presented. The high efficiency is obtained by...
详细信息
ISBN:
(纸本)9781467321877
An efficient algorithm for analysis and optimization of rotationally symmetric reflector antennas, possibly including 3D support structures and/or waveguide components, is presented. The high efficiency is obtained by using a domain-decomposition approach where each region of space is characterized independently using a generalized admittance matrix description. The admittance matrices are obtained using circular mode-matching, higher-order 3D MoM, or a newly developed higher-order MoM for bodies of revolution (BoR-MoM). The antenna performance is rigorously evaluated by cascading the admittance matrices, which subsequently allows computation of surface currents or fields in all subdomains. The new algorithm allows fast and accurate analysis and optimization of rotationally symmetric reflectors, even in cases that would normally require a time-consuming 3D solution due to the lack of rotational symmetry.
暂无评论