The Alzheimer's disease (AD) is the most prevalent neurodegenerative brain disease worldwide. The neuroimaging based computer-aided diagnosis (CAD) of AD has attracted much attention. Therefore, machine learning p...
详细信息
The Alzheimer's disease (AD) is the most prevalent neurodegenerative brain disease worldwide. The neuroimaging based computer-aided diagnosis (CAD) of AD has attracted much attention. Therefore, machine learning plays an important role in it. However, it is usually a time-costing task to label a large-scale data in clinical practice, which results in the neuroimaging dataset for AD is a small sample size. The semi-supervised learning (SSL), which has the ability to employ the easily acquired unlabeled data to assist insufficient labeled data for improving learning performance, have attracted considerable interest for AD classification. The co-training classification is a representative paradigm of disagreement-based SSL method. It trains alternately to maximize the mutual agreement on two preferably independent views (feature subsets) of the unlabeled data. Since the magnetic resonance imaging (MRI) and positron emission tomography (PET) are the commonly used neuroimaging techniques for AD diagnosis, co-training method has the potential to perform semi-supervised classification of AD with both MRI and PET data. In this paper, we propose the co-training based SSL with MRI and PET images for classification of AD with the mild cognitive impairment (MCI) data as the unlabeled samples. We compare the co-training method with other supervised and semi-supervised classification method on the ADNI dataset. The experimental results indicate that co-training method with both MRI and PET data achieves best performance, which suggest that it has the potential to be applied to neuroimaging based CAD for AD.
The process of staircases detection and recognition is complex for blinds. Therefore, an intelligent and real time system is required to help them. In this paper, we investigate using only one ultrasonic sensor and fe...
详细信息
The process of staircases detection and recognition is complex for blinds. Therefore, an intelligent and real time system is required to help them. In this paper, we investigate using only one ultrasonic sensor and few samples with small size to represent floor and staircases. The performance of such system depend on object representation, data modeling and finally classification algorithm. A simple wave analysis have shown that frequency components are the most affected in stair case context. Accordingly, we have used frequency representation of ultrasonic signal, namely the smoothed periodogram. Then, we model model several extracted features based on Masson possibility approach. Finally, similarity measure is used in the classification algorithm. A training process is under taken on a local database of 500 signal simples is used. An accuracy rate of 94% has been achieved.
Acquisition is an essential building block in almost any real life dataprocessing. In acquisition it is key target to keep aliasing low often dictates the use of a complex analog anti-aliasing filter. In case of mult...
详细信息
Acquisition is an essential building block in almost any real life dataprocessing. In acquisition it is key target to keep aliasing low often dictates the use of a complex analog anti-aliasing filter. In case of multichannel system, each channel must be fitted with a separate anti-aliasing filter as such filter cannot be multiplexed but this exercise becomes expensive. In this paper the cost optimized structure for multiple channel data acquisition system is proposed, the cost optimization achieved by using multistage decimation. The MATLAB simulation has been used to evaluate proposed model and in result cost reduced by 53.27% when (M=16*2) has been used instead of M=32. The cost has been further reduced by 69% for combination of M=(8*4), Key targets are filter length reduction and aliasing. Finite-impulse-response (FIR) has been used for implementation of idea.
This paper presents an algorithm based on the multiple model approach for tracking highly maneuverable targets. The proposed method aims at propagating a low number of model hypotheses while still yielding low estimat...
详细信息
This paper presents an algorithm based on the multiple model approach for tracking highly maneuverable targets. The proposed method aims at propagating a low number of model hypotheses while still yielding low estimation errors compared to other multiple model algorithms. Merging and pruning techniques are used to keep very small sets of model sequences matched with the target dynamics. Simulation results show that the proposed method is able to attain superior performance during periods of no maneuver compared to the Interacting Multiple Model algorithm as well as when abrupt changes take place in the target movement.
The future of far-infrared observations rests on our capacity to reach sub-arcsecond angular resolution around 100 mu m, in order to achieve a significant advance with respect to our current capabilities. Furthermore,...
详细信息
ISBN:
(纸本)9780819496119
The future of far-infrared observations rests on our capacity to reach sub-arcsecond angular resolution around 100 mu m, in order to achieve a significant advance with respect to our current capabilities. Furthermore, by reaching this angular resolution we can bridge the gap between capacities offered by the JWST in the near infrared and those allowed by ALMA in the submillimeter, and thus benefit from similar resolving capacities over the whole wavelength range where interstellar dust radiates and where key atomic and molecular transitions are found. In an accompanying paper, 1 we present a concept of a deployable annular telescope, named TALC for Thinned Aperture Light Collector, reaching 20m in diameter. Being annular, this telescope features a main beam width equivalent to that of a 27m telescope, i.e. an angular resolution of 0.92" at 100 mu m. In this paper we focus on the science case of such a telescope as well on the aspects of unconventional dataprocessing that come with this unconventional optical configuration. The principal science cases of TALC revolve around its imaging capacities, that allow resolving the Kuiper belt in extra-solar planetary systems, or the filamentary scale in star forming clouds all the way to the Galactic Center, or the Narrow Line Region in Active Galactic Nuclei of the Local Group, or breaking the confusion limit to resolve the Cosmic Infrared Background. Equipping this telescope with detectors capable of imaging polarimetry offers as well the extremely interesting perspective to study the influence of the magnetic field in structuring the interstellar medium. We will then present simulations of the optical performance of such a telescope. The main feature of an annular telescope is the small amount of energy contained in the main beam, around 30% for the studied configuration, and the presence of bright diffraction rings. Using simulated point spread functions for realistic broad-band filters, we study the observing performanc
Recently finite-resolution digital receivers have been proposed for weighted-transmitted reference ultra-wideband (WTR-UWB) systems. However, employing a finite-resolution receiver severely degrades the performance du...
详细信息
Recently finite-resolution digital receivers have been proposed for weighted-transmitted reference ultra-wideband (WTR-UWB) systems. However, employing a finite-resolution receiver severely degrades the performance due to a sever quantization noise. In this paper, two novel iterative algorithms are proposed to recover the reference pulse based on the detected data pulses and employing nonlinear signalprocessing while preventing the error propagation. The first algorithm exploits a new signalprocessing approach, referred to as polarity invariant square law technique. In this approach the reconstructed reference pulse is first squared and multiplied by its sign and then is used to demodulate the monobit data pulses. Hence, it can be assumed that in this technique data pulses have been recovered from monobit quantization noise as well. In the second algorithm, the mentioned reconstructed reference pulse is cubed and then is used for demodulation. Simulation results illustrate that employing these two novel signal-processing methods highly mitigates the contribution of the small (noise dominant) samples to the decision statistic. The performance of the proposed algorithms is evaluated at a data rate of 23.78 Mbps over in-vehicle channels, taking into account noise, inter-symbol interference (ISI), and inter-block interference (IBI). The obtained simulation results demonstrate that the proposed algorithms enhance the performance about 1.7 dB as compared to the existing algorithms. As a result, applying these algorithms would be a promising approach to implement high performance monobit transmitted-reference (TR) receivers.
MicroRNAs (miRNAs) are a class of small non-coding RNAs of 22 nucleotides which normally function as negative regulators of target mRNA expression at the posttran-scriptional level. miRNAs play a role for one or more ...
详细信息
MicroRNAs (miRNAs) are a class of small non-coding RNAs of 22 nucleotides which normally function as negative regulators of target mRNA expression at the posttran-scriptional level. miRNAs play a role for one or more target genes by suppressing in processes as growth, differentiation, proliferation and cell death. Recent evidence has shown that miRNA mutations or mis-expression correlate with various human cancers and indicates that miRNAs can function as tumour suppressors and oncogenes. MicroRNAs have been shown to repress the expression of important cancer-related genes and might prove useful in the diagnosis and treatment of cancer. In this study, hierarchical microRNA clusters are obtained through microarray expression data in order to analyze the microRNA prostate cancer relationships. Clustering results are evaluated by their biological relevance. It is seen that such approach can be useful in detectitn relationships between microRNAs and diseases.
Human detection offers many advantages in applications of search and rescue, smart environments, and security. Infrared, acoustic, vibration/seismic and visual sensors have been often used in human detection and recog...
详细信息
Human detection offers many advantages in applications of search and rescue, smart environments, and security. Infrared, acoustic, vibration/seismic and visual sensors have been often used in human detection and recognition systems. Radar offers unique advantages for sensing humans, such as remote operation during virtually all weather conditions, increased range, and better coverage. However, radar systems are typically very expensive and physically large. The BumbleBee radar, in contrast to most radars, is a low power pulse Doppler radar that is about the size of a business card. Moreover, it is a radar that can be integrated into indoor wireless sensor networks. In this work, the application of BumbleBee radar to human activity recognition by computing the human micro-Doppler signature is examined. Humans are complex targets that are capable of many motions. Every part of the human causes different reflection and every motion of the human has its unique micro-doppler signatures. The differences in micro-Doppler data of activities such as walking, running, and crawling that is gathered from low-cost, low-power radar is discussed.
Recently, semi-supervised sparse feature selection, which can exploit the large number unlabeled data and small number labeled data simultaneously, has placed an important role in web image annotation. However, most o...
详细信息
Recently, semi-supervised sparse feature selection, which can exploit the large number unlabeled data and small number labeled data simultaneously, has placed an important role in web image annotation. However, most of the semi-supervised feature selection methods are developed for single-view data, which can not reveal and leverage the correlated and complemental information between different views. Recently, multi-view learning has obtained much research attention, so we apply multi-view learning into semi-supervised sparse feature selection and propose a multi-view semi-supervised sparse feature selection method based on graph Laplacian, namely Multi-view Laplacian Sparse Feature Selection (MLSFS) in this paper. MLSFS can realize sparse feature selection by utilizing the correlated and complemental information between different views. A simple iterative method is proposed to solve the objective function of MLSFS. We apply our algorithm into image annotation and conduct experiments on two web image datasets. The results show that the proposed multi-view method outperforms the single-view methods.
A new signalprocessing method using a single vector hydrophone is proposed for solving the problem of azimuth angle estimation for multiple targets based on a small aperture underwater *** method extends the aperture...
详细信息
A new signalprocessing method using a single vector hydrophone is proposed for solving the problem of azimuth angle estimation for multiple targets based on a small aperture underwater *** method extends the aperture from a single vector hydrophone into a half wavelength distance uniform linear array by decomposing the time-domain sample data from different channels of the *** extended array has a narrower space beam pattern than a single vector *** azimuths of spatial multiple targets are estimated by using the appended array snapshots under the condition of broadband or narrow band *** new method is robust because there is no need to correct the array *** analysis and computer simulations show that,the new algorithm has the ability to distinguish two incoherent targets with either narrow band or broadband signals in an isotropic noise *** algorithm provides a non-biased estimate with a high signal-to-noise ratio.
暂无评论