Matching Items (107)

Filtering by

Clear all filters

149867-Thumbnail Image.png

Incorporating auditory models in speech/audio applications

Description

Following the success in incorporating perceptual models in audio coding algorithms, their application in other speech/audio processing systems is expanding. In general, all perceptual speech/audio processing algorithms involve minimization of an objective function that directly/indirectly incorporates properties of human perception.

Following the success in incorporating perceptual models in audio coding algorithms, their application in other speech/audio processing systems is expanding. In general, all perceptual speech/audio processing algorithms involve minimization of an objective function that directly/indirectly incorporates properties of human perception. This dissertation primarily investigates the problems associated with directly embedding an auditory model in the objective function formulation and proposes possible solutions to overcome high complexity issues for use in real-time speech/audio algorithms. Specific problems addressed in this dissertation include: 1) the development of approximate but computationally efficient auditory model implementations that are consistent with the principles of psychoacoustics, 2) the development of a mapping scheme that allows synthesizing a time/frequency domain representation from its equivalent auditory model output. The first problem is aimed at addressing the high computational complexity involved in solving perceptual objective functions that require repeated application of auditory model for evaluation of different candidate solutions. In this dissertation, a frequency pruning and a detector pruning algorithm is developed that efficiently implements the various auditory model stages. The performance of the pruned model is compared to that of the original auditory model for different types of test signals in the SQAM database. Experimental results indicate only a 4-7% relative error in loudness while attaining up to 80-90 % reduction in computational complexity. Similarly, a hybrid algorithm is developed specifically for use with sinusoidal signals and employs the proposed auditory pattern combining technique together with a look-up table to store representative auditory patterns. The second problem obtains an estimate of the auditory representation that minimizes a perceptual objective function and transforms the auditory pattern back to its equivalent time/frequency representation. This avoids the repeated application of auditory model stages to test different candidate time/frequency vectors in minimizing perceptual objective functions. In this dissertation, a constrained mapping scheme is developed by linearizing certain auditory model stages that ensures obtaining a time/frequency mapping corresponding to the estimated auditory representation. This paradigm was successfully incorporated in a perceptual speech enhancement algorithm and a sinusoidal component selection task.

Contributors

Agent

Created

Date Created
2011

152307-Thumbnail Image.png

Adaptive learning and unsupervised clustering of immune responses using microarray random sequence peptides

Description

Immunosignaturing is a medical test for assessing the health status of a patient by applying microarrays of random sequence peptides to determine the patient's immune fingerprint by associating antibodies from a biological sample to immune responses. The immunosignature measurements can

Immunosignaturing is a medical test for assessing the health status of a patient by applying microarrays of random sequence peptides to determine the patient's immune fingerprint by associating antibodies from a biological sample to immune responses. The immunosignature measurements can potentially provide pre-symptomatic diagnosis for infectious diseases or detection of biological threats. Currently, traditional bioinformatics tools, such as data mining classification algorithms, are used to process the large amount of peptide microarray data. However, these methods generally require training data and do not adapt to changing immune conditions or additional patient information. This work proposes advanced processing techniques to improve the classification and identification of single and multiple underlying immune response states embedded in immunosignatures, making it possible to detect both known and previously unknown diseases or biothreat agents. Novel adaptive learning methodologies for un- supervised and semi-supervised clustering integrated with immunosignature feature extraction approaches are proposed. The techniques are based on extracting novel stochastic features from microarray binding intensities and use Dirichlet process Gaussian mixture models to adaptively cluster the immunosignatures in the feature space. This learning-while-clustering approach allows continuous discovery of antibody activity by adaptively detecting new disease states, with limited a priori disease or patient information. A beta process factor analysis model to determine underlying patient immune responses is also proposed to further improve the adaptive clustering performance by formatting new relationships between patients and antibody activity. In order to extend the clustering methods for diagnosing multiple states in a patient, the adaptive hierarchical Dirichlet process is integrated with modified beta process factor analysis latent feature modeling to identify relationships between patients and infectious agents. The use of Bayesian nonparametric adaptive learning techniques allows for further clustering if additional patient data is received. Significant improvements in feature identification and immune response clustering are demonstrated using samples from patients with different diseases.

Contributors

Agent

Created

Date Created
2013

151215-Thumbnail Image.png

Energy and quality-aware multimedia signal processing

Description

Today's mobile devices have to support computation-intensive multimedia applications with a limited energy budget. In this dissertation, we present architecture level and algorithm-level techniques that reduce energy consumption of these devices with minimal impact on system quality. First, we present

Today's mobile devices have to support computation-intensive multimedia applications with a limited energy budget. In this dissertation, we present architecture level and algorithm-level techniques that reduce energy consumption of these devices with minimal impact on system quality. First, we present novel techniques to mitigate the effects of SRAM memory failures in JPEG2000 implementations operating in scaled voltages. We investigate error control coding schemes and propose an unequal error protection scheme tailored for JPEG2000 that reduces overhead without affecting the performance. Furthermore, we propose algorithm-specific techniques for error compensation that exploit the fact that in JPEG2000 the discrete wavelet transform outputs have larger values for low frequency subband coefficients and smaller values for high frequency subband coefficients. Next, we present use of voltage overscaling to reduce the data-path power consumption of JPEG codecs. We propose an algorithm-specific technique which exploits the characteristics of the quantized coefficients after zig-zag scan to mitigate errors introduced by aggressive voltage scaling. Third, we investigate the effect of reducing dynamic range for datapath energy reduction. We analyze the effect of truncation error and propose a scheme that estimates the mean value of the truncation error during the pre-computation stage and compensates for this error. Such a scheme is very effective for reducing the noise power in applications that are dominated by additions and multiplications such as FIR filter and transform computation. We also present a novel sum of absolute difference (SAD) scheme that is based on most significant bit truncation. The proposed scheme exploits the fact that most of the absolute difference (AD) calculations result in small values, and most of the large AD values do not contribute to the SAD values of the blocks that are selected. Such a scheme is highly effective in reducing the energy consumption of motion estimation and intra-prediction kernels in video codecs. Finally, we present several hybrid energy-saving techniques based on combination of voltage scaling, computation reduction and dynamic range reduction that further reduce the energy consumption while keeping the performance degradation very low. For instance, a combination of computation reduction and dynamic range reduction for Discrete Cosine Transform shows on average, 33% to 46% reduction in energy consumption while incurring only 0.5dB to 1.5dB loss in PSNR.

Contributors

Agent

Created

Date Created
2012

152455-Thumbnail Image.png

On the ordering of communication channels

Description

This dissertation introduces stochastic ordering of instantaneous channel powers of fading channels as a general method to compare the performance of a communication system over two different channels, even when a closed-form expression for the metric may not be available.

This dissertation introduces stochastic ordering of instantaneous channel powers of fading channels as a general method to compare the performance of a communication system over two different channels, even when a closed-form expression for the metric may not be available. Such a comparison is with respect to a variety of performance metrics such as error rates, outage probability and ergodic capacity, which share common mathematical properties such as monotonicity, convexity or complete monotonicity. Complete monotonicity of a metric, such as the symbol error rate, in conjunction with the stochastic Laplace transform order between two fading channels implies the ordering of the two channels with respect to the metric. While it has been established previously that certain modulation schemes have convex symbol error rates, there is no study of the complete monotonicity of the same, which helps in establishing stronger channel ordering results. Toward this goal, the current research proves for the first time, that all 1-dimensional and 2-dimensional modulations have completely monotone symbol error rates. Furthermore, it is shown that the frequently used parametric fading distributions for modeling line of sight exhibit a monotonicity in the line of sight parameter with respect to the Laplace transform order. While the Laplace transform order can also be used to order fading distributions based on the ergodic capacity, there exist several distributions which are not Laplace transform ordered, although they have ordered ergodic capacities. To address this gap, a new stochastic order called the ergodic capacity order has been proposed herein, which can be used to compare channels based on the ergodic capacity. Using stochastic orders, average performance of systems involving multiple random variables are compared over two different channels. These systems include diversity combining schemes, relay networks, and signal detection over fading channels with non-Gaussian additive noise. This research also addresses the problem of unifying fading distributions. This unification is based on infinite divisibility, which subsumes almost all known fading distributions, and provides simplified expressions for performance metrics, in addition to enabling stochastic ordering.

Contributors

Agent

Created

Date Created
2014

152344-Thumbnail Image.png

Adaptive methods within a sequential Bayesian approach for structural health monitoring

Description

Structural integrity is an important characteristic of performance for critical components used in applications such as aeronautics, materials, construction and transportation. When appraising the structural integrity of these components, evaluation methods must be accurate. In addition to possessing capability to

Structural integrity is an important characteristic of performance for critical components used in applications such as aeronautics, materials, construction and transportation. When appraising the structural integrity of these components, evaluation methods must be accurate. In addition to possessing capability to perform damage detection, the ability to monitor the level of damage over time can provide extremely useful information in assessing the operational worthiness of a structure and in determining whether the structure should be repaired or removed from service. In this work, a sequential Bayesian approach with active sensing is employed for monitoring crack growth within fatigue-loaded materials. The monitoring approach is based on predicting crack damage state dynamics and modeling crack length observations. Since fatigue loading of a structural component can change while in service, an interacting multiple model technique is employed to estimate probabilities of different loading modes and incorporate this information in the crack length estimation problem. For the observation model, features are obtained from regions of high signal energy in the time-frequency plane and modeled for each crack length damage condition. Although this observation model approach exhibits high classification accuracy, the resolution characteristics can change depending upon the extent of the damage. Therefore, several different transmission waveforms and receiver sensors are considered to create multiple modes for making observations of crack damage. Resolution characteristics of the different observation modes are assessed using a predicted mean squared error criterion and observations are obtained using the predicted, optimal observation modes based on these characteristics. Calculation of the predicted mean square error metric can be computationally intensive, especially if performed in real time, and an approximation method is proposed. With this approach, the real time computational burden is decreased significantly and the number of possible observation modes can be increased. Using sensor measurements from real experiments, the overall sequential Bayesian estimation approach, with the adaptive capability of varying the state dynamics and observation modes, is demonstrated for tracking crack damage.

Contributors

Agent

Created

Date Created
2013

153878-Thumbnail Image.png

New passive methodology for power cable monitoring and fault location

Description

The utilization of power cables is increasing with the development of renewable energy and the maintenance replacement of old overhead power lines. Therefore, effective monitoring and accurate fault location for power cables are very important for the sake of a

The utilization of power cables is increasing with the development of renewable energy and the maintenance replacement of old overhead power lines. Therefore, effective monitoring and accurate fault location for power cables are very important for the sake of a stable power supply.

The recent technologies for power cable diagnosis and temperature monitoring system are described including their intrinsic limitations for cable health assessment. Power cable fault location methods are reviewed with two main categories: off-line and on-line data based methods.

As a diagnostic and fault location approach, a new passive methodology is introduced. This methodology is based on analyzing the resonant frequencies of the transfer function between the input and output of the power cable system. The equivalent pi model is applied to the resonant frequency calculation for the selected underground power cable transmission system.

The characteristics of the resonant frequencies are studied by analytical derivations and PSCAD simulations. It is found that the variation of load magnitudes and change of positive power factors (i.e., inductive loads) do not affect resonant frequencies significantly, but there is considerable movement of resonant frequencies under change of negative power factors (i.e., capacitive loads).

Power cable fault conditions introduce new resonant frequencies in accordance with fault positions. Similar behaviors of the resonant frequencies are shown in a transformer (TR) connected power cable system with frequency shifts caused by the TR impedance.

The resonant frequencies can be extracted by frequency analysis of power signals and the inherent noise in these signals plays a key role to measure the resonant frequencies. Window functions provide an effective tool for improving resonant frequency discernment. The frequency analysis is implemented on noise laden PSCAD simulation signals and it reveals identical resonant frequency characteristics with theoretical studies.

Finally, the noise levels of real voltage and current signals, which are acquired from an operating power plant, are estimated and the resonant frequencies are extracted by applying window functions, and these results prove that the resonant frequency can be used as an assessment for the internal changes in power cable parameters such as defects and faults.

Contributors

Agent

Created

Date Created
2015

154392-Thumbnail Image.png

A Case Study of Executive Stock Compensation Design for The State-Owned Firms

Description

Executive compensation design involving equity shares has been widely used in Europe, the United States and other developed countries where the capital markets are relatively mature. In China, due to the differences in industries, ownership structure, stages of enterprise development,

Executive compensation design involving equity shares has been widely used in Europe, the United States and other developed countries where the capital markets are relatively mature. In China, due to the differences in industries, ownership structure, stages of enterprise development, constraints faced by the firms, the executive compensation design using equity shares tends to vary accordingly. For the state-owned companies, the situations are more complex than others. This complexity has not been a focus of the past literature, particularly on the compensation contract design and its subsequent implementation. Based on Coase contract theorem, agency theory and human capital theory, I examined how different state-owned firms vary in their approaches on managerial stock compensation design using a case study approach. The thesis concludes with a summary of major findings and a discussion of policy implications.

Contributors

Agent

Created

Date Created
2016

154410-Thumbnail Image.png

Investment Style and Performance Attribution Analysis on Chinese A Share Market

Description

With the fast development of Chinese capital market, an increasing number of institutions and retail investors invest through professional managers. The key to evaluating investment manager’s skill and performance persistence largely lies in portfolio style research and attribution analysis.

With the fast development of Chinese capital market, an increasing number of institutions and retail investors invest through professional managers. The key to evaluating investment manager’s skill and performance persistence largely lies in portfolio style research and attribution analysis.

The current dissertation takes advantage of a unique dataset, uncover hidden investment style and trading behavior, understanding their source of excess returns, and establishing a more comprehensive methodology for evaluating portfolio performance and manager skills.

The dissertation focuses on quantitative analysis. Highlights three most important aspects. Investment style determines the systematic returns and risks of any portfolio, and can be assessed ex-ante; Transaction can be observed and modified during the investment process; and return attribution can be implemented to evaluate portfolio (managers), ex-post. Hence, these three elements make up a comprehensive and logical investment process.

Investment style is probably the most important factor in determining portfolio returns. However, Chinese investment managers are under constant pressure to follow the market trend and shift style accordingly. Therefore, accurately identifying and predicting each manager’s investment style proves critically valuable.

In addition, transaction data probably provides the most reliable source of information in observing and evaluating an investment manager’s style and strategy, in the middle of the investment process.

Despite the efficacy of traditional return attribution methodology, there are clear limitations. The current study proposes a novel return attribution methodology, by synthesizing major portfolio strategy components, such as risk exposure adjustment, sector rotation, stock selection, altogether. Our novel methodology reveals that investment managers do not obtain much abnormal returns through risk exposure adjustment or sector rotation. Instead, Chinese investment managers seem to enjoy most of their excess returns through stock selection.

In addition, we find several interesting patterns in Chinese A-share market: 1). There is a negative relationship between asset under management (AUM) and investment performance, beyond certain AUM threshold; 2). There are limited benefits from style switching in the long run; 3). Many investment managers use CSI 300 component stocks as portfolio ballast and speculate with CSI500 and Medium-and-Small board component stocks for excess returns; 4). There is no systematic negative relationship between portfolio turnover and investment performance; despite negative relationship within certain sub-samples and sectors; 5). It is plausible to construct out-performing portfolios with style index funds and ETFs.

Contributors

Agent

Created

Date Created
2016

154374-Thumbnail Image.png

A Case Study of Credit Risk Analysis and Modeling for SMEs -In an Internet Finance Setting

Description

In the last two years, China’s booming of Internet Finance Platform made significant impacts on three dimensions. Compared with the conventional market, Internet Finance is asserted to open a revolutionary pathway of lending where by small and mid-sized companies may

In the last two years, China’s booming of Internet Finance Platform made significant impacts on three dimensions. Compared with the conventional market, Internet Finance is asserted to open a revolutionary pathway of lending where by small and mid-sized companies may overcome the financing dilemma on credit accessibility and high cost. In other words, Internet Finance is hyped to be able to reduce information asymmetry, enhance allocation efficiency of resources, and promote product and process innovations for the financial institutions. However, the core essence of Internet Finance rests on risk assessment and control – a fundamental element applies to all forms of financing. Most current practice of internet finance on risk assessment and control remains unchanged from the mindset of traditional banking practices for small and medium sized firms. Hence, the same problems persisted and may only become even worse under the internet finance platform if no innovations take place.

In this thesis, the author proposed and tested a credit risk assessment model using data analytics techniques through an in-depth cases study with actual transaction data. Specifically, based on the 30,000 observations collected from actual transactional data from small and medium size firms of China’s home furnishing industry. The preliminary results are promising in spite of the limitations. The thesis concludes with the findings of relevance to improve the current practices and suggests areas of future research.

Contributors

Agent

Created

Date Created
2016

154377-Thumbnail Image.png

Research on Factors Influencing Individual’s Behavior of Energy Management

Description

With the rapid rise of distributed generation, Internet of Things, and mobile Internet, both U.S. and European smart home manufacturers have developed energy management solutions for individual usage. These applications help people manage their energy consumption more efficiently. Domestic manufacturers

With the rapid rise of distributed generation, Internet of Things, and mobile Internet, both U.S. and European smart home manufacturers have developed energy management solutions for individual usage. These applications help people manage their energy consumption more efficiently. Domestic manufacturers have also launched similar products.

This paper focuses on the factors influencing Energy Management Behaviour (EMB) at the individual level. By reviewing academic literature, conducting surveys in Beijing, Shanghai and Guangzhou, the author builds an integrated behavioural energy management model of the Chinese energy consumers. This paper takes the vague term of EMB and redefines it as a function of two separate behavioural concepts: Energy Management Intention (EMI), and the traditional Energy Saving Intention (ESI).

Secondly, the author conducts statistical analyses on these two behavioural concepts. EMI is the main driver behind an individual’s EMB. EMI is affected by Behavioural Attitudes, Subjective Norms, and Perceived Behavioural Control (PBC). Among these three key factors, PBC exerts the strongest influence. This implies that the promotion of the energy management concept is mainly driven by good application user experience (UX). The traditional ESI also demonstrates positive influence on EMB, but its impact is weaker than the impacts arising under EMI’s three factors. In other words, the government and manufacturers may not be able to change an individual's energy management behaviour if they rely solely on their traditional promotion strategies. In addition, the study finds that the government may achieve better promotional results by launching subsidies to the manufacturers of these kinds of applications and smart appliances.

Contributors

Agent

Created

Date Created
2016