Engineering Commons^™

Use Of Principal Component Analysis With Linear Predictive Features In Developing A Blind Snr Estimation System, Matthew James Marbach

Theses and Dissertations

Signal-to-noise ratio is an important concept in electrical communications, as it is a measurable ratio between a given transmitted signal and the inherent background noise of a transmission channel. Currently signal-to-noise ratio testing is primarily performed by using an intrusive method of comparing a corrupted signal to the original signal and giving it a score based on the comparison. However, this technique is inefficient and often impossible for practical use because it requires the original signal for comparison. A speech signal's characteristics and properties could be used to develop a non-intrusive method for determining SNR, or a method that does …

Data Fusion Of Complementary Information From Parietal And Occipital Event Related Potentials For Early Diagnosis Of Alzheimer's Disease, Nicholas Stepenosky

Theses and Dissertations

The number of the elderly population affected by Alzheimer's disease is rapidly rising. The need to find an accurate, inexpensive, and non-intrusive procedure that can be made available to community healthcare providers for the early diagnosis of Alzheimer's disease is becoming an increasingly urgent public health concern. Several recent studies have looked at analyzing electroencephalogram signals through the use of many signal processing techniques. While their methods show great promise, the final outcome of these studies has been largely inconclusive. The inherent difficulty of the problem may be the cause of this outcome, but most likely it is due to …

Random Feature Subspace Ensemble Based Approaches For The Analysis Of Data With Missing Features, Hussein Syed Mohammed

Theses and Dissertations

Missing data in real world applications is not an uncommon occurrence. It is not unusual for training, validation or field data to have missing features in some (or even all) of their instances, as bad sensors, failed pixels, malfunctioning equipment, unexpected noise causing signal saturation, data corruption, and so on, are all familiar scenarios in many practical applications.

In this thesis, the feasibility of an ensemble of classifiers trained on a feature subset space is investigated as an effective and practical solution for the missing feature problem. Two ensemble of classifiers approach motivated by the Random Subspace Method are proposed …

Boosted Ensemble Algorithm Strategically Trained For The Incremental Learning Of Unbalanced Data, Michael David Muhlbaier

Theses and Dissertations

Many pattern classification problems require a solution that needs to be incrementally updated over a period of time. Incremental learning problems are often complicated by the appearance of new concept classes and unbalanced cardinality in training data. The purpose of this research is to develop an algorithm capable of incrementally learning from severely unbalanced data. This work introduces three novel ensemble based algorithms derived from the incremental learning algorithm, Learn++. Learn++.NC is designed specifically for incrementally learning New Classes through dynamically adjusting the combination weights of the classifiers' decisions. Learn++.UD handles Unbalanced Data through class-conditional voting …

An Architecture For Intelligent Health Assessment Enabled Ieee 1451 Compliant Smart Sensors, Donald Albert Nickles

Theses and Dissertations

As systems become increasingly complex and costly, potential failure mechanisms and indicators are numerous and difficult to identify, while the cost of loss is very expensive - human lives, replacement units, and impacts to national security. In order to ensure the safety and long-term reliability of vehicles, structures, and devices attention must be directed toward the assessment and management of system health. System health is the key component that links data, information, and knowledge to action. Integrated Systems Health Management (ISHM) doctrine calls for comprehensive real-time health assessment and management of systems where the distillation of raw data into information …

Multiresolution Wavelet Analysis Of Event-Related Eeg Potentials Using Ensemble Of Classifier Data Fusion Techniques For Early Diagnosis Of Alzheimer's Disease, Apostolos Topalis

Theses and Dissertations

The recent advances and knowledge in medicine and nutrition have greatly improved our average life expectancy. An unfortunate consequence of this longer life span, however, is a dramatic increase in the number of individuals suffering from dementia, and more specifically, from Alzheimer's disease (AD). Furthermore, AD remains under-diagnosed and under-treated until its more severe stages due to lack of standard diagnostic tools available to community clinics. A search for biomarkers that will allow early diagnosis of the disease is therefore necessary to develop effective medical treatments. Such a biomarker should be non-invasive, simple to obtain, safe, inexpensive, accurate, and most …

A "Divide-And-Conquer" Strategy For Nde Signal Inversion In Gas Transmission Pipelines, Justin Gary Bram

Theses and Dissertations

Signal inversion in nondestructive evaluation (NDE) applications is a critical step before remediation decisions are made. The accuracy and confidence of the signal inversion results therefore play a key role in evaluating the effectiveness of the NDE procedure. Conventional NDE signal inversion algorithms that employ artificial neural networks treat all geometric regions of the NDE signal equally. Consequently, when the inversion algorithm is presented with input data that is significantly different from the training data, the performance of the network deteriorates significantly. This thesis presents a superior alternative for NDE signal inversion. Different geometric regions of the NDE signature are …