Search Results (1 - 25 of 200 Results)

Sort By  
Sort Dir
 
Results per page  

Khajotia, Burzin K.CASE BASED REASONING – TAYLOR SERIES MODEL TO PREDICT CORROSION RATE IN OIL AND GAS WELLS AND PIPELINES
Master of Science (MS), Ohio University, 2007, Industrial and Manufacturing Systems Engineering (Engineering)

Corrosion rate prediction involves developing a predictive model that provides a realistic estimate, utilizing common operational parameters, existing lab/field data, and theoretical models. The novel Case-based Reasoning – Taylor Series (CBR-TS) model for corrosion prediction developed in this thesis, takes knowledge from existing field cases and uses CBR techniques and Taylor series to predict corrosion rates for new fields having similar parameters. The model predicts corrosion in three steps: case search (selection of similar cases), case ranking (by using Taylor series expansion), and case revision (by using correction factor from mechanistic or semi-empirical model). The model is implemented as a prototype and verified on a large hypothetical case base and a small field database with real data.

Committee:

Dusan Sormaz (Advisor)

Keywords:

CBR; Case Based Prediction; Case Based Reasoning; Corrosion Prediction; Corrosion Rate Prediction

Ghareeb, AhmedData mining for University of Dayton campus buildings to predict future demand
Master of Science (M.S.), University of Dayton, 2017, Mechanical Engineering
The ability to forecast demand for large facilities will be increasingly important as real-time power pricing scenarios become increasingly present. Accurate prediction will inform data-driven power shedding to reduce energy costs most effectively with minimal sacrifice of comfort. A number of previous researchers have researched this topic, achieving results with varying amount of success. This study looks to forecast demand for a university complex of buildings, subject to the unique occupancy variation of such institutions. Specifically addressed is the use of academic institutional data associated with temporal enrollment and the academic calendar. As well, it addresses use of demand data in all buildings in an effort to more accurately predict this aggregate demand of the university. A data mining based approach based upon a Random Forest regression tree algorithm is used to develop the forecast model. The mean absolute percentage error (MAPE) value associated with the model applied to a validation set of data is on the order of 2.21 % based upon actual weather data. Using forecasted weather data, the MAPE increases to approximately 6.65 % in predicted day-ahead demand.

Committee:

Kevin Hallinan (Committee Chair); Andrew Chiasson (Committee Member); Zhongmei Yao (Committee Member)

Subjects:

Artificial Intelligence; Climate Change; Energy; Engineering; Environmental Engineering; Mechanical Engineering; Statistics

Keywords:

Data mining; energy prediction; energy demand; energy demand forecasting; energy; prediction; forecasting; modeling;

Angus, G. D.Prediction of academic and clinical success at the Ohio State University School of Allied Medical Professions /
Doctor of Philosophy, The Ohio State University, 1972, Graduate School

Committee:

Not Provided (Other)

Subjects:

Health Sciences

Keywords:

Prediction of occupational success;Prediction of scholastic success

Pattakos, GregoryPredicting Length of Stay and Non-Home Discharge: A Novel Approach to Reduce Wasted Resources after Cardiac Surgery
Master of Sciences, Case Western Reserve University, 2011, Clinical Research

Objectives: We sought to understand factors associated with non-home discharge and postoperative length of stay and to develop validated prediction models for each.

Methods: Logistic regression analysis of non-home discharge and linear regression of length of stay were performed using preoperative patient demographics, noncardiac comorbidities, cardiac symptoms and comorbidities, and planned procedure.

Results: Factors available for preadmission discharge planning associated with increased postoperative length of stay included female gender, lower socioeconomic position, elevated BMI, and aortic surgery. Preoperative prediction of postoperative length of stay was poor (R=0.66 for development cohort and 0.59 for validation cohort). Factors associated with non-home discharge included age, greater home-to-hospital distance, emergency status, and aortic surgery. High predictability was achieved (C-statistic=.81 for development cohort and 0.83 for validation cohort).

Conclusions: Preadmission prediction of non-home discharge is accurate. Application of this model will allow preadmission planning, possibly leading to reduction of wasted resources after cardiac surgery.

Committee:

Ralph O'Brien, PhD (Committee Chair); Eugene Blackstone, MD (Committee Member); Edward Nowicki, MD (Committee Member)

Subjects:

Surgery

Keywords:

length of stay; cardiac surgery; non-home discharge; prediction; resource use; prediction model; SNF; rehabilitation; heart surgery;

Wiley, Matthew T.Machine Learning for Diabetes Decision Support
Master of Science (MS), Ohio University, 2011, Computer Science (Engineering and Technology)

This thesis presents work in machine learning that enhances and expands the scope of the 4 Diabetes Support System (4DSS). The 4DSS is a decision support system designed to assist patients and physicians with the challenge of managing Type 1 diabetes (T1DM). The objective of the 4DSS is to detect problems in diabetes management and to recommend therapeutic changes to correct these detected problems. This thesis contributes three advances: (1) preprocessing noisy data, preparatory to applying machine learning algorithms; (2) enhancing the automated detection of excessive glycemic variability, a serious problem for patients with diabetes; and (3) predicting patient blood glucose levels, in order to preemptively detect and avoid potential health problems.

In this work, the Continuous Glucose Monitoring (CGM) data is smoothed using cubic spline smoothing with extra weight on fingersticks and local optima. This data preprocessing improves the accuracy of problem detection and blood glucose prediction. Previous work in classifying glycemic variability using a naıve Bayes classifier obtained an accuracy of only 87.1%. Using smoothed CGM data and a rich set of domain independent pattern recognition features to train multilayer perceptrons and support vector machines, a best accuracy of 93.8% has now been obtained. This machine learning classifier improves the ability to detect excessive glycemic variability, an important indicator of risk for diabetic complications.

Accurately predicting blood glucose levels could enhance patient safety by giving patients time to intervene before problems occur. Support Vector Regression (SVR) and AutoRegressive Integrated Moving Average (ARIMA) models were built and tested on data from ten T1DM patients. This resulted in a best Root Mean Square Error (RMSE) of 4.5 mg/dl for 30 minute predictions and 17.4 mg/dl for 60 minute predictions. Clarke Error Grid Analysis (CEGA) showed that 99% of 30 minute predictions and 90% of 60 minute predictions fell within 20% of target, CEGA's most accurate range.

Committee:

Cynthia Marling, PhD (Advisor); Razvan Bunescu, PhD (Committee Member); Frank Schwartz, MD (Committee Member); Jundong Liu, PhD (Committee Member)

Subjects:

Computer Science; Medicine

Keywords:

diabetes; machine learning; glycemic variability; time series prediction; decision support; glucose prediction

Attal, AsadullahDevelopment of Neural Network Models for Prediction of Highway Construction Cost and Project Duration
Master of Science (MS), Ohio University, 2010, Civil Engineering (Engineering and Technology)

Early understanding of construction cost and time represents a critical factor of a feasibility study in the early design phase of a project. All parties involved in the construction of a project; owners, contractors, and services companies are in need of reliable information about the cost and time in the early stages of the project. Therefore, this research work attempts to develop a consistent model of forecasting early design construction cost of highway and the project’s duration. A wide review of the literature of the cost and project duration determined the significance of reliable methods to predict construction cost and project’s duration. At this time, researchers suffer lack of information in the early phases of project to identify and analyze parameters and their relationship related to duration of projects. However, there have been numerous attempts to develop improved models for construction cost prediction of highway based on different techniques.

This research aims to develop construction cost and duration prediction models of highway based on statistical analysis. Consequently, the statistical techniques used in this research work represent Artificial Neural Networks (ANN) and step wise regression analysis to identify the influential parameters and forecast the early design phase of highway construction cost and duration. The input data used to develop the mathematical models were compiled and maintained by the Virginia Department of Transportation .The data used in these modeling were extracted from two sources within VDOT: Data Warehouse Management Information Portal (DWMIP) and Project Cost Estimating System (PCES)”. The parametric stage data were maintained in Project Cost Estimating System (PCES) by VDOT. In addition, for the identification of effective parameters used in these models, two separate techniques were used; sorting and identifying the effective parameters used in traditional techniques, the trial and elimination method of ANNs, and sensitivity analysis. Consequently, the chosen parameters were analyzed by two distinct statistical techniques: linear regression analysis and non linear ANN. Also, the given data were classified and analyzed for full depth section and improvement of highway and each section was classified as a full or condensed model. The same classification and analytical procedure was used for both the highway’s cost prediction and the projects’ duration prediction.

As the result of effective parameters identification and prediction models analysis the ANN outcome represented higher accuracy and reliability than linear regression analysis. Also, the identification of influential parameters represents a crucial impact on the future investigations.

Committee:

Mehmet Tatari, PhD (Advisor); Byong Cheol Kim, PhD (Committee Member); Munir Nazzal, PhD (Committee Member); Martin J Mohlenkamp, PhD (Committee Member)

Subjects:

Civil Engineering

Keywords:

Highway Cost Prediction; Highway Construction Duration Prediction; Artificial Neural Networks; Regresssion; Influential Parameters

Zhao, ShuangFORWARD AND BACKWARD EXTENDED PRONY (FBEP) METHOD WITH APPLICATIONS TO POWER SYSTEM SMALL-SIGNAL STABILITY
Doctor of Philosophy, Case Western Reserve University, 2017, EECS - System and Control Engineering
We introduce the “Forward and Backward Extended Prony” (FBEP) method that identifies the parameters of complex exponential signals using a new strategy for finding true pole locations. The performance of the proposed method is investigated theoretically using statistical analysis and experimentally by simulation. Initial validation is accomplished using time series data without additive noise and with the help of singular value decomposition (SVD), the advantages of this method in accurately identifying both growing and decaying modes in moderate noise is then demonstrated by adding noise to the time series data with different signal-to-noise ratios (SNRs). The FBEP method is compared with the TLS-Prony method and the subspace-based methods by illustrating the Mean Squared Errors (MSEs) of the frequency and damping factor estimates given by each method with comparisons to the corresponding Cramer-Rao (CR) bounds. The computational time of FBEP and the subspace-based methods is compared. The performance of the FBEP method using the pseudoinverse and Total Least Squares (TLS) approaches below the threshold SNR is also studied. The FBEP method can be applied in cases where the poles of some modes are known a priori in complex exponential signals. The FBEP method is applied to power system small-signal stability analysis. Its effectiveness in identifying system eigenvalues from the output signal is validated by experiments on a test system model. Using a four-machine-two-area power system model the identification of the dominant modes contained in oscillatory signals given by the FBEP method is compared with that given by the SVD-TLS method, the Prony-SR method and Trudnowski’s algorithm at different SNR levels. The computational cost of the FBEP method, the Prony-SR method, and Trudnowski’s algorithm is evaluated. The results from multi-signal analysis and sliding window analysis using the FBEP method are also presented.

Committee:

Kenneth Loparo (Committee Chair); Vira Chankong (Committee Member); Marc Buchner (Committee Member); Richard Kolacinski (Committee Member); Mingguo Hong (Committee Member)

Subjects:

Engineering

Keywords:

Prony; Forward linear prediction; Backward linear prediction; Complex exponentials; Power system; Low-frequency oscillation; Eigenvalue;

Iommi, MorganUsing the Integrative Model of Behavior Prediction to Understand Factors Influencing Graduate Teaching Assistants’ Teaching Development Attendance
Doctor of Philosophy, The Ohio State University, 2017, Communication
This dissertation proposes a modified version of the Integrative Model of Behavior Prediction (Fishbein, 2000) to understand motivations affecting Ohio State University Graduate Teaching Assistants (GTAs)’ behavioral intentions for attending teaching development workshops at their university’s center for teaching (UCAT). Current Ohio State GTAs (N = 139) were surveyed to explore how attitudes, norms, efficacy, and anticipated emotions affect their behavioral intentions. The study found support for some elements of the modified model, including the additions of response efficacy and anticipated emotions. Anticipated emotions were found to work as a mediator for the main variables’ effects on behavioral intention. Support was not found for the interaction effect of injunctive and descriptive norms. The study also found that the distal variables of previous teaching experience and familiarity with their university’s center for teaching affected behavioral intentions to attend teaching development workshops at the GTAs’ center for teaching. Implications for behavior prediction research and teaching development implementation and marketing are discussed.

Committee:

Shelly Hovick, PhD (Advisor); Roselyn Lee-Won, PhD (Advisor); Nancy Rhodes, PhD (Committee Member)

Subjects:

Communication

Keywords:

Graduate Teaching Assistants, Teaching Development, Behavior Prediction, Integrative Model of Behavior Prediction

Pappada, Scott MichaelPrediction of Glucose for Enhancement of Treatment and Outcome: A Neural Network Model Approach
Doctor of Philosophy in Engineering, University of Toledo, 2010, Bioengineering

Critical care (e.g. trauma and cardiothoracic surgical) and diabetic patients are prone to variability in glucose concentration on a daily basis. Hypoglycemic and hyperglycemic glucose values in these patient populations have been associated with decreased patient outcomes. In diabetic patients, persistently elevated glucose values are associated with development of long term complications such as, but not limited to retinopathy, neuropathy, and nephropathy. In the critical care patient population, elevated glucose has been correlated to increases in mortality, length of stay in the intensive care unit (ICU), and morbidities. The maintenance of tight glycemic control in these patients without severe hypoglycemia or glycemic variability appears to improve outcomes in these patients.

Various factors are associated with future glycemic excursions such as, but not limited to: lifestyle/activities (e.g. sleep-wake cycles), emotional factors (e.g. stress), nutritional intake, medication dosages, and ICU medical records (in critical care patients). In the field of diabetes research, models for prediction of glucose and/or models used to maintain tight glycemic control have been the focus of research. In the critical care patient population, very little research into development of such models has been completed to date.

Multiple factors affect or are indicators of future glucose concentration. A suitable modeling technique needs to incorporate the effect of such factors for accurate prediction of glucose. A modeling technique well suited for this task is a neural network model.A neural network is an adapative modeling technique, which learns and updates model parameters based on determining patterns/trends existent in input data.This adapative capability, makes neural network modeling well suited for prediction of glucose where multiple factors impact future glycemic excursions.

This dissertation summarizes the development and optimization of various neural network model architectures for the real-time prediction of glucose in diabetic and critical care patients. Neural network models were configured to predict glucose using prediction horizons >60 minutes, which have not been attained in many predictive models to date. The performance of the neural network model is assessed via determination of overall model error, percentage of glycemic extremes predicted, and clinical acceptability of model predictions as determined via Clarke Error Grid Analysis.

Committee:

Brent Cameron, PhD (Advisor); Ronald Fournier, PhD (Committee Member); Thomas Papadimos, MD (Committee Member); Marilyn Borst, MD (Committee Member); William Olorunto, MD (Committee Member)

Subjects:

Bioinformatics; Computer Science; Engineering

Keywords:

prediction of glucose; diabetes; critical care; neural network; closed loop insulin delivery; glucose; real-time prediction;

Minca, Kristen KathleenUsing Soil Nutrient Tests and 1M HNO3 to Predict Total and Bioaccessible Pb in Urban Soils
Master of Science, The Ohio State University, 2012, Environment and Natural Resources
Urban redevelopment has created vacant land in many old industrial cities. Use of urban land for food production, including gardening, involves human exposure to soil. This can be a human health issue if the soil contains historical contaminants such as Pb. Most urban soils are not tested for Pb because of the high costs associated with sampling and laboratory analysis of soil contaminants. However, soil testing for plant nutrients is inexpensive and routinely performed for agricultural soils used for food production. The objectives of this study are (1) to compare the ability of 1M HNO3, Mehlich 3, and Modified Morgan soil tests to predict total Pb in urban soils and (2) to evaluate the ability of 1M HNO3, Mehlich 3, and Modified Morgan soil tests to predict bioaccessible Pb in urban soils. Total and bioaccessible Pb was determined from 65 urban residential vacant lots being considered for urban gardens and food production in Cleveland, OH. Extractable Pb was determined using common soil nutrient test methods Mehlich 3 and Modified Morgan extraction, and a 1M HNO3 extraction. The results of this study show Mehlich 3, 1M HNO3, and Modified Morgan were strongly correlated with total and bioaccessible Pb. Both Modified Morgan and 1M HNO3 are not commonly used soil tests. Most commercial and university soil testing labs use Mehlich 3 to measure available plant nutrients. These laboratories are more likely to use their existing Mehlich 3 soil test to estimate Pb. They are unlikely to add a new soil test (Modified Morgan or 1M HNO3) solely for estimating Pb. The Mehlich 3 soil test could be used as a screening tool to not only estimate total Pb (slope 1.83, r2 = 0.959) but also to estimate bioaccessible Pb when using RBALP at pH 1.5 (slope 1.47, r2 =0.965) and RBALP at pH 2.5 (slope 0.92, r2 = 0.943). Because Mehlich 3 is currently the most widely used test to evaluate available plant nutrients and because it is relatively inexpensive (< $15) it can easily be adopted by soil testing laboratories to screen samples for Pb. Our results show that total Pb can be conservatively estimated by the following equation Total Pb (mg kg-1) = Mehlich 3 Pb (mg kg-1) x 2.

Committee:

Nicholas Basta (Advisor); Brian Slater (Committee Member); Jeff Sharp (Committee Member)

Subjects:

Agriculture; Environmental Science; Public Health; Soil Sciences

Keywords:

Mehlich-3; urban agriculture; lead; bioaccessible; soil test; prediction

GURUMURTHY, MADHUSUDHANA ROBUST DECISION-AIDED MIMO CHANNEL ESTIMATION SCHEME
MS, University of Cincinnati, 2006, Engineering : Electrical Engineering
In this thesis, we present a decision-aided channel estimator for multiple-input multiple-output (MIMO) Rayleigh fading channels. The scheme presented does away with the conventional block fading channel model by making continual updates to the channel estimate using previous decisions. We tailor the WLMS algorithm proposed in [1] which takes advantage of knowledge of the nature of channel fading for prediction, to suit MIMO channel needs. We then analyze the symbol error rate, channel prediction error floor achieved and also test the robustness of the proposed algorithm. The results reveal superior performance of our scheme to previous works in MIMO channel prediction, in all of the areas mentioned above and robustness to decision errors as well.

Committee:

Dr. James Caffery (Advisor)

Keywords:

MIMO; channel; estimation; prediction;

Scott-Emuakpor, Onome EjaroDevelopment of a novel energy-based method for multi-axial fatigue strength assessment
Doctor of Philosophy, The Ohio State University, 2007, Mechanical Engineering
An accelerated method for determining the fatigue stress versus cycle life (S-N) behavior of isotropic materials is developed for prediction of axial (tension-compression), bending, shear, and multi-axial fatigue life at various stress ratios. The framework for this accelerated method was developed in accordance with a previous understanding of a strain energy and fatigue life correlation, which states: the total strain energy dissipated during a monotonic fracture and a cyclic process is the same material property, where each can be determined by measuring the area underneath the monotonic true stress-strain curve and the area within a hysteresis loop, respectively. The developed framework consists of the following six elements: (1) New experimental procedures used to acquire more sufficient uniaxial and multi-axial test results than conventional methods, (2) an analytical representation for the effect of the stress gradient through the fatigue zone, thus providing capability for bending fatigue prediction, (3) the effect of mean stress on fatigue life for tension/compression and bending, (4) development of an improved energy-based prediction criterion for shear loading at various stress ratios, (5) fatigue life prediction for materials experiencing the endurance limit phenomenon, and (6) the development of a multi-axial fatigue life prediction method. Validation of this accelerated fatigue life determination framework is achieved based on comparison with numerous experimental results acquired from Aluminum 6061-T6 and Titanium 6Al-4V. The results of the comparison are extremely encouraging, thus providing justification that the future direction for the strain-energy based fatigue life prediction method is very promising.

Committee:

Mo-How Shen (Advisor)

Subjects:

Engineering, Mechanical

Keywords:

fatigue; multiaxial; prediction; energy; failure; bending; uniaxial

Velissariou, PanagiotisDevelopment of a Coastal Prediction System That Incorporates Full 3D Wave-Current Interactions on the Mean Flow and the Scalar Transport With Initial Application to the Lake Michigan Turbidity Plume
Doctor of Philosophy, The Ohio State University, 2009, Civil Engineering

The present work focuses on the development of a Modular Multi-Component Coastal Ocean Prediction System (mmcops) that incorporates the full 3D wave-current interactions for a better representation of the entrainment and transport mechanics in complex deep and shallow water coastal environments. The system incorporates wind, temperature and atmospheric pressure forcing that drive the circulation, wave, sediment and bottom boundary layer model components.

The effects of the wind generated surface waves on the water column and bottom layer dynamics are parametrized by the inclusion of the Stokes drift, and the wave radiation stress terms that quantify the excess of mass and momentum flux produced by the waves. Coupled wave-hydrodynamic models traditionally incorporate the radiation stress terms only into the vertically integrated momentum. Considering the fact that currents are 3D structures, the vertical variation of the radiation stress should be also considered. In the present work the 3D momentum equations are re-derived to include the full 3D impact of the radiation stresses on the currents.

As a preliminary test, the system is applied to Lake Michigan with a twofold purpose: a to conduct an initial testing of the model prognostic variables with and without the effect of the waves; and b to develop a methodology required to answer whether the annually observed Spring turbidity nearshore plume in Southern Lake Michigan is transporting material from its origin in one continuous transport mode or as generated by a series of local deposition, resuspension and transport activities. To this end data collected during the EEGLE project are fully analyzed; shoreline erosion rates and texture of the eroded material were collected from various sources and via various methods and are presented for 34 shoreline segments in a uniform format; an Eulerian Particle Tracking formulation that identifies the source and origin of the various particle sizes within the sediment plume is presented; and a conceptual and computational set up of the control volumes or sediment plume sources/origins required for a detailed study of the Spring turbidity plume is developed.

Committee:

Keith Bedford, W (Advisor); Carolyn Merry, J (Committee Member); Gil Bohrer (Committee Member)

Subjects:

Civil Engineering; Geophysics; Ocean Engineering; Oceanography

Keywords:

prediction system; surface waves; 3D wave effects; 3D wave radiation stress; wave current boundary layer; wave breaking; sediment transport; shoreline erosion; equation of state; heat flux; Lake Michigan; turbidity plume; MAROBS

Rao, Dhananjai M.Study of Dynamic Component Substitutions
PhD, University of Cincinnati, 2003, Engineering : Computer Science and Engineering
High fidelity, high resolution models of systems need to be simulated for conducting in-depth studies of different scenarios and to ensure that crucial scalability issues do not dominate during validation of simulation results. However, simulation of large, high resolution models is a time consuming task. Consequently, the models are statically (i.e., before simulation commences) abstracted to improve performance of the simulations and minimize analysis overheads. However, abstraction improves performance by trading resolution, and possibly the fidelity, of the simulations –which defeats the purpose of studying high resolution models and magnifies the problems of validation! An alternate approach to improve the overall efficiency of simulation studies is to dynamically (i.e., during simulation) change the resolution of the model. Accordingly, this study proposes and explores the use of a novel methodology called Dynamic Component Substitution to enable dynamic changes to the resolution of the model. In DCS, a set of components (called a module) are dynamically substituted by a functionally equivalent component; thereby changing the resolution of a model without impacting the overall validity of the model. DCS improves the overall efficiency of simulations by enabling dynamic tradeoffs between several modeling and simulation related parameters. Therefore, it is crucial to use ideal sequences of component substitutions to ensure optimal simulation performance and meet the analysis requirements. However, identifying optimal sequences of DCS, particularly in parallel simulation environments is a complex task. Consequently, to ease effective use of DCS, a DCS algebra (i.e., a mathematical framework) along with a DCS Performance Prediction Methodology (DCSPPM) has been developed. This study empirically explores the practical applicability and effectiveness of DCS by applying it to several models from a variety of domains. The design and development of a modeling and parallel simulation environment to enable effective use of DCS is discussed. The issues involved in the implementation of DCSPPM are presented. The study also presents an empirical evaluation of the accuracy of the estimates generated by DCSPPM. The results from these studies indicate that DCS can significantly reduce simulation time in a predictable manner without impacting the overall validity of the simulation study.

Committee:

Dr. Philip A. Wilsey (Advisor)

Keywords:

modeling; multi-resolution, variable-resolution; parallel simulation; performance prediction; web-based modeling & simulation

Kelly, John KipEstimation of Behavioral Thresholds in Normal Hearing Listeners Using Auditory Steady State Responses
Doctor of Philosophy, The Ohio State University, 2009, Speech and Hearing Science

The ability to obtain frequency specific information regarding a patient’s hearing sensitivity in an objective manner allows the evaluation of patient populations who cannot be tested through traditional behavioral methods. One method for obtaining this information is the auditory steady state response (ASSR). ASSR permits the testing of multiple carrier frequencies simultaneously and both ears simultaneously, unlike the auditory brainstem response (ABR). ASSR replaces subjective examiner interpretation of the response with statistical analyses not subject to the variability of human observers.

Unlike ABR which has been in use for decades and utilizes relatively consistent stimuli and test protocols, the ASSR has only been in widespread clinical use for the past 6-8 years and consequently does not have the same level of standardization as ABR. ASSR can be elicited by a variety of stimulus types including but not limited to: (1) Sinusoidal amplitude modulated (SAM) tones, (2) Frequency modulated (FM) tones, (3) Mixed modulation (MM) tones, and (4) Toneburst (TB) trains. The ASSR response is found in the frequency domain at the frequency of modulation and is frequently differentiated from unrelated neural activity using an F-statistic to determine if the amplitude of the line spectra at the modulation frequency is statistically different from the surround physiologic noise.

The current study sought to evaluate several common stimuli used in ASSR testing to determine if a more recently introduced stimulus (TB) emerges as a more appropriate stimulus for generating the response. Response detection and collection parameters were standardized so that any differences seen could be attributed to the stimulus. Both behavioral and ASSR thresholds were measured using SAM, MM, and TB stimuli in ten young adults with normal hearing (≤ 15 dB HL from 250-8000 Hz). Comparisons were then made between stimulus types to determine which stimuli could best predict a behavioral response for a pure tone matching the carrier frequency.

The results of the current study indicate that the MM and TB stimuli provide lower ASSR thresholds than do SAM stimuli and that a regression model provides the most accurate estimates of behavioral threshold. The thresholds for an individually presented TB were consistently lower than for a TB at the same frequency that was presented in the multiple simultaneous paradigm (four simultaneous carrier frequencies presented to the ear). However the threshold predictions based on the two measurements were similar so little accuracy in prediction is lost by using multiple simultaneously presented tonebursts. The current study shows that while ASSR can provide reasonable estimates of hearing sensitivity when the mean data are examined for any given individual the accuracy of prediction can vary greatly.

Committee:

Lawrence Feth, Ph.D. (Committee Chair); Christina Roup, Ph.D. (Committee Member); Ashok Krishnamurthy, Ph.D. (Committee Member)

Subjects:

Audiology

Keywords:

Auditory Steady State Response; ASSR; Hearing Threshold Prediction

Wertz, John NicholasAn Energy-Based Experimental-Analytical Torsional Fatigue Life-Prediction Method
Master of Science, The Ohio State University, 2010, Aero/Astro Engineering
An energy-based cycle-dependent fatigue life prediction framework for the calculation of torsional fatigue life and remaining life has been developed. The framework for this fatigue prediction method is developed in accordance with previously developed energy-based axial and bending fatigue life prediction approaches, which state: the total strain energy density accumulated during both a monotonic fracture event and cyclic processes is the same material property, where each can be determined by measuring the area beneath the monotonic true stress-strain curve and the area within a hysteresis loop, respectively. The energy-based fatigue life prediction framework is composed of the following entities: (1) the development of a shear fatigue testing procedure capable of assessing cyclic plastic strain energy density accumulation in a pure shear stress state and (2) the incorporation of an energy-based fatigue life calculation scheme to determine the remaining fatigue life of in-service gas turbine materials subjected to pure shear fatigue. Validation of the improved theory was attempted but failed due to undesired axial loading occurring during testing. Future work was proposed to address the issues.

Committee:

Herman Shen, PhD (Advisor); Jack McNamara, PhD (Committee Member); Tommy George, PhD (Other); Onome Scott-Emuakpor, PhD (Other)

Subjects:

Aerospace Materials; Engineering; Experiments; Mechanical Engineering; Mechanics

Keywords:

energy-based torsional fatigue life-prediction shear method

Gray, Anna R.Creation of a Modified Equation to Predict VO2 on a Cycle Ergometer
Master of Science (MS), Ohio University, 2012, Exercise Physiology-Research (Health Sciences and Professions)
INTRODUCTION: Prediction equations are inexpensive, easy and commonly used to estimate the oxygen consumption (VO2) and caloric expenditure of physical activity, but some of these equations have limitations, including a stated limit of accuracy of the cycling equation up to a work rate (WR) of 200 watts. This is not favorable to trained individuals who can cycle at WR much higher than 200 watts. PURPOSE: To create a VO2 prediction equation for cycling WR below and above 200 watts. METHODS: Twenty nine participants qualified for this study by achieving a maximal WR (WRmax) of at least 300 watts during a maximal graded exercise test (GXT). These individuals then completed two submaximal exercise trials (SXT). During the SXT, steady state VO2 was collected for each WR and was analyzed by linear regression. RESULTS: The 29 participants were 20-54 years old, 1.80 ¿¿¿¿ 0.00 meters tall, weighing 78.29 ¿¿¿¿ 0.44 kg with a body fat of 9.14 ¿¿¿¿ 0.21% and a WRmax of 361.13 ¿¿¿¿ 2.06 watts. According to the analysis of this sample, the data suggests VO2 (ml/kg/min) = {10.941ml/kg/min x (WR (watts) / body weight (kg))} + 4.522 (ml/kg/min) with an R2 of 0.96. CONCLUSION: The regression equation created from this data to predict VO2 is useful to estimate the oxygen consumption and caloric expenditure of trained individuals who cycle below and above 200 watts.

Committee:

Michael Kushnick, PhD (Committee Chair); Michael Clevidence, Mr. (Committee Member); Roger Gilders, PhD (Committee Member); Cheryl Howe, PhD (Committee Member); Gordon Brooks, PhD (Committee Member)

Subjects:

Health Sciences; Kinesiology; Physical Education; Physiology

Keywords:

VO2; oxygen consumption; prediction equation; trained cyclists; aerobic athletes; cycling

Lufaso, Michael WaynePerovskite Synthesis and Analysis Using Structure Prediction Diagnostic Software
Doctor of Philosophy, The Ohio State University, 2002, Chemistry

A software program, SPuDS (Structure Prediction Diagnostic Software) was developed to predict the crystal structures of perovskites, including those distorted by tilting of symmetric and Jahn-Teller distorted octahedra. Only the composition and oxidation states of the ions were required as inputs. Rigid octahedra were maintained in electronically symmetric octahedra, while distortions of the octahedral bond lengths were utilized in Jahn-Teller distorted compositions. SPuDS calculates the optimal structure in Glazer tilt systems most often observed experimentally and for tilt systems with interesting multiple A-cation environments. Structure optimization occurs by distorting the structure to minimize the global instability index. Location of the A-site cation is chosen to maximize the symmetry of its coordination environment.

SPuDS has been employed in a number of useful applications, including use of predicted structural information to estimate physical properties of both hypothetical compositions and those materials for which accurate structural data is unavailable, as a guide for exploratory synthetic efforts, as a starting model for Rietveld refinements in the course of structurally characterizing materials, and for extracting the effects of octahedral tilting distortions from other structural distortion mechanisms. Structural predictions were made for a variety of compositions with single and multiple octahedral cations and compared with a previously determined structures to illustrate the accuracy of this approach.

Synthesis and structural refinement from x-ray and neutron powder diffraction of Ca2MnMO6 (M = Nb, Sb, Ru) and Sr2MnMO6 (M=Nb, Sb) is reported. Various types of cooperative Jahn-Teller distortions (CJTD) that occur in ternary (AMX3) and double (A2MM′O6) perovskites are reviewed. Interplay between CJTD’s, octahedral tilting and cation order is systematically examined in A2MM′O6 perovskites, where M is an active Jahn-Teller (J-T) ion (Mn3+ or Cu2+) and M′ is an octahedrally symmetric cation.

Perovskites with a+a+a+ octahedral tilting and a single octahedral cation exhibit interesting dielectric properties. SPuDS was also used to examine the prospects for synthesizing new compounds with multiple A-cation coordination geometries. Perovskites with octahedral tilting (a+a+a+) coupled with A-cation and rock-salt octahedral cation ordering were predicted to exist. High-pressure high-temperature synthesis, characterization of the structures and dielectric properties of CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12, CaCu3Ga2Nb2O12 and CaCu3Ga2Ta2O12 is reported.

Committee:

Patrick Woodward (Advisor)

Subjects:

Chemistry, Inorganic

Keywords:

perovskite; structure prediction; structure analysis; Jahn-Teller

Wells, Brian JayPredicting and Preventing Colorectal Cancer
Doctor of Philosophy, Case Western Reserve University, 2012, Epidemiology and Biostatistics

Colorectal cancer (CRC) is the third leading cause of cancer mortality among both men and women in the United States despite prevention efforts. A better tool for predicting colorectal cancer risk may help identify high risk patients who could benefit from chemopreventive medication, but the best statistical method for reducing survival models in large datasets is controvertible.

The first project compared several variable selection methodologies in their ability to produce parsimonious, accurate prediction models in large datasets. The second investigation created a model for predicting colorectal cancer risk using a large prospective dataset of almost 200,000 participants. The final project was a cost-effectiveness analysis that explored the potential benefits of sulindac-difluoromethylornithine as an adjunct to colonoscopy for the prevention of CRC.

A modified version of Forward Stepwise Regression based on model discrimination (Forward Stepwise C-Statistic) produced slightly more accurate prediction models than traditional forms of stepwise regression based on the F Statistic or Harrell’s Stepdown. New, gender-specific colorectal cancer risk calculators were created using Forward Stepwise C-statistic and achieved bias corrected C-Statistics (95% CI) of 0.69 (0.673-0.698) and 0.68 (0.668-0.694) in men and women, respectively. Age was the overriding risk factor for both men and women. The cost effectiveness analysis revealed that screening colonoscopy is both less expensive and results in more quality adjusted life years than any of the following strategies: no prevention, colonoscopy + chemoprevention, and chemoprevention alone. This finding persisted across one-way sensitivity analyses as well high risk patient scenarios.

Forward Stepwise C-Statistic can produce reduced statistical models that are slightly more accurate than traditional stepwise regression or Harrell’s Stepdown, but the slight gain in accuracy may not be worth the additional computational burden in some instances. Head to head comparisons between the new colorectal cancer risk calculator and existing calculators are necessary to prove that the new model is better. The overriding influence of age in the new model suggests that current strategies to screen patients based on age may not be unreasonable. The results demonstrate that chemoprevention with sulindac-difluoromethylornithine should not be considered for patients without a hereditary cancer syndrome.

Committee:

Leila Jackson, PhD (Committee Chair); Siran Koroukian, PhD (Committee Member); Michael Kattan, PhD (Committee Member); Mendel Singer, PhD (Committee Member); Gregory Cooper, MD (Committee Member)

Subjects:

Epidemiology

Keywords:

colorectal cancer; risk prediction; cost-effectivness analysis; chemoprevention; variable selection

Holm, Jeannette E.Collision Prediction and Prevention in a Simultaneous Multi-User Immersive Virtual Environment
Master of Computer Science, Miami University, 2012, Computer Science & Software Engineering
Immersive virtual environments allow users to explore the virtual world by physically walking. Because users are fully immersed in the virtual environment, they have no visual reference to their physical surroundings. Since users only see the virtual world, collisions between users who are simultaneously in the tracking area are inevitable. Immersive virtual environments are thus limited in their ability to support concurrent users. A technique called redirected walking has been used to alter user paths by imperceptibly rotating the world users see, guiding them away from tracking area boundaries. Using this technique, users can walk naturally through virtual worlds for miles, unaware of the physical boundaries of the tracking area. This thesis describes the design and implementation of an extension to existing redirected walking algorithms. This extension is designed to predict and prevent collisions. If a collision between users is predicted, redirected walking techniques are applied to steer them away from one another.

Committee:

Dr. Eric Bachmann, PhD (Advisor); Dr. Eric Hodgson, PhD (Committee Member); Dr. Michael Zmuda, PhD (Committee Member)

Subjects:

Computer Science

Keywords:

redirected walking; virtual reality; immersive; virtual environments; multi-user; collision prediction; collision prevention; collision avoidance

KUNAPULI, UDAYKUMARA STUDY OF SWAP CACHE BASED PREFETCHING TO IMPROVE VITUAL MEMORY PERFORMANCE
MS, University of Cincinnati, 2002, Engineering : Computer Engineering
With dramatic increase in processor speeds over the last decade, disk latency has become a critical issue in computer systems performance. Disks, being mechanical devices, are orders of magnitude slower than the processor or physical memory. Most Virtual Memory(VM) systems use disk as secondary storage for idle data pages of an application. The working set of pages is kept in memory. When a page requested by the processor is not present in memory, it results in a page fault. On a page fault, the Operating System brings the requested page from the disk into memory. Thus the performance of Virtual Memory systems depends on disk performance. In this project, we aim to reduce the effect of disks on Virtual Memory performance compared to the traditional demand paging system. We study novel techniques of page grouping and prefetching to improve Virtual Memory system performance. We group pages, evicted from memory at about the same time, into a single large block. On a page fault, we prefetch the entire block along with the faulting page. We implement this grouping and prefetching scheme with a swap cache. The swap cache combines a group of pages, evicted from memory, into a superblock. Superblock is the basic unit of I/O operation during paging and swapping. During a disk read, the entire superblock that has the required page is read from the disk directly into memory. We prefetch all pages with memory eviction locality in a single disk read. From this study, we find that swap cache based prefetching significantly reduces the number of read accesses to the disk. Our simulations show that the number of read accesses to the disk reduced by at least 12% for all the six SPEC 2000 benchmark applications used in this study. For some applications, the number of read accesses reduced by as much as 90%. We also find improvement in Virtual Memory I/O performance of many SPEC 2000 benchmark applications. With the swap cache, Virtual Memory performance of five of the six SPEC 2000 benchmark applications improved by at least 25%, with some improving up to 88%.

Committee:

Dr.Yiming Hu (Advisor)

Keywords:

virtual memory; page fault; swap cache; prediction; operating systems

Suttman, Alexander K.Lithium Ion Battery Aging Experiments and Algorithm Development for Life Estimation
Master of Science, The Ohio State University, 2011, Mechanical Engineering

Battery lifespan is one of the largest considerations when designing battery packs for electrified vehicles. Even during vehicle operation, it is essential to monitor the progression of a battery health as it degrades and predict battery life. This thesis presents a preliminary severity factor analysis based on available experimental data and details the development of an algorithm for predicting, while in operation, the remaining life of a battery based on the growth of internal resistance.

Nine lithium ion batteries were systematically aged through severe aging protocols spanning multiple C-rates (2C, 4C and 8C), low ranges of SOC (0-10, 0-20 and 0-30%), and elevated temperature (55 deg C). Their internal resistance was continuously calculated at each sharp current transition, and these values were filtered and processed. Severity factors were calculated for each battery by determining the average rate of resistance growth over a battery life and a preliminary analysis of these factors was carried out. A resistance growth dynamic model was developed to identify rates of resistance growth on a local basis as resistance values were collected. These local rates of resistance growth were then used to calculate predicted future rates of resistance growth, which were in turn used to predict remaining life.

The life prediction algorithm produced continuously updated predictions of remaining battery life that proved relatively accurate for cases of constant battery aging conditions. This computationally simple algorithm could be implemented onboard an electrified vehicle to provide estimates of remaining battery life based on resistance growth. This methodology can in principle be readily extended to track capacity degradation as well, provided that a feasible capacity estimator can be developed on the basis of vehicle measurements.

Committee:

Yann Guezennec (Advisor); Giorgio Rizzoni (Committee Member); Simona Onori (Committee Member)

Subjects:

Automotive Engineering; Electrical Engineering; Mechanical Engineering

Keywords:

lithium; lithium ion; battery; aging; life; prediction; li-ion; resistance; algorithm; degradation; electric; hev; phev; hybrid

Swaminathan, KarthikeyanEnhanced prediction of Phosphorylation and Disorder in Proteins
PhD, University of Cincinnati, 2009, Engineering : Biomedical Engineering
Over the years, many predictors of structural and functional properties of proteins have beendeveloped on the basis that this information is encoded in the protein sequence. The fact that excellent prediction techniques are available has put the spotlight on the representation of the protein sequence i.e. the input features to these techniques. In this study, we focus on the structural properties of flexible regions and assess three specific conformational flexibility parameters (i) RSA Confidence that is readily available from our in-house secondary structure and solvent accessibility predictor, SABLE (ii) X-ray structure derived B-factors, which we enhance and (iii) NMR structure derived solvent accessibility standard deviations (SASDs), which is a feature we propose here. In the case of B-factors, a combination of PSSMs and real valued SS/RSA predictions, including RSA Confidence had been used to enhance its prediction. In the case of NMR SASDs we have presented a novel predictor that exploits the same feature set as B-factors. In each case, we have developed an epsilon-support vector regression (e-SVR) model towards this. To our knowledge, the use of SASDs as input features and the development of a predictor for the same is novel. It has also been shown that it might be easier to predict SASDs as compared to B-factors. The three flexibility parameters were then applied to the prediction of conformational disorder as well as prediction of phosphorylation. In the case of disorder, we have shown through cross-validation on our training set, that the addition of RSA confidence to the input feature space, improves its prediction significantly and is further improved with the addition of B-factor and SASD predictions developed in this study. All the three parameters were then included in the final predictor, which gave the top performance on the CASP8 data set in terms of average accuracy and CASP weighted scores. Even the removal of an easy target (T0500) did not dislodge our predictor from giving the top weighted score. In the case of phosphorylation, we have shown that the addition of real-valued SS and RSA predictions significantly improve the prediction as evaluated by cross-validation on the training set and is further improved by addition of B-factors and SASDs together. Additionally, we present a comparison of one- and two-class support vector machines (SVMs) as applied to the prediction of phosphorylation. In the prediction of phosphorylation, the methods typically employ a two-class classification approach with the limitation that the set of negative examples used for training may include some sites that are simply unknown to be phosphorylated. While one-class classification techniques have been considered in the past as a solution to this problem, their performance has not been systematically compared to two-class techniques. In this study, we developed and compared one- and two-class SVM based predictors for several commonly used sets of attributes. [These predictors are being made available at http://sable.cchmc.org/] Keywords: phosphorylation, protein disorder, prediction, b-factors, solvent accessibility.

Committee:

Jaroslaw Meller, PhD (Committee Chair); Marepalli Rao, PhD (Committee Member); Mario Medvedovic, PhD (Committee Member)

Subjects:

Biomedical Research

Keywords:

phosphorylation;protein disorder;b-factors;solvent accessibility;prediction

Krishnamurthy, RevathyKnowledge Enabled Location Prediction of Twitter Users
Master of Science (MS), Wright State University, 2015, Computer Science
As the popularity of online social networking sites such as Twitter and Facebook continues to rise, the volume of textual content generated on the web is increasing rapidly. The mining of user generated content in social media has proven effective in domains ranging from personalization and recommendation systems to crisis management. These applications stand to be further enhanced by incorporating information about the geo-position of social media users in their analysis. Due to privacy concerns, users are largely reluctant to share their location information. As a consequence of this, researchers have focused on automatic inferencing of location information from the contents of a user’s tweets. Existing approaches are purely data-driven and require large training data sets of geotagged tweets. Furthermore, these approaches rely solely on social media features or probabilistic language models and fail to capture the underlying semantics of the tweets. In this thesis, we propose a novel knowledge based approach that does not require any training data. Our approach uses Wikipedia, a crowd sourced knowledge base, to extract entities that are relevant to a location. We refer to these entities as local entities. Additionally, we score the relevance of each local entity with respect to the city, using the Wikipedia Hyperlink Graph. We predict the most likely location of the user by matching the scored entities of a city and the entities mentioned by users in their tweets. We evaluate our approach on a publicly available data set consisting of 5119 Twitter users across continental United States and show comparable accuracy to the state-of-the-art approaches. Our results demonstrate the ability to pinpoint the location of a Twitter user to a state and a city using Wikipedia, without needing to train a probabilistic model.

Committee:

Amit Sheth, Ph.D. (Advisor); Krishnaprasad Thirunarayan, Ph.D. (Committee Member); Derek Doran, Ph.D. (Committee Member)

Subjects:

Computer Science

Keywords:

Wikipedia, Twitter, Location Prediction, Semantics

Henry, Emily BrookeStochastic Modeling of Geometric Mistuning and Application to Fleet Response Prediction
Master of Science in Engineering (MSEgr), Wright State University, 2014, Mechanical Engineering
An improved spatial statistical approach and probabilistic prediction method for mistuned integrally bladed rotors is proposed and validated with a large population of rotors. Prior work utilized blade-alone principal component analysis to model spatial variation arising from geometric deviations contributing to forced response mistuning amplification. Often, these studies considered a single rotor measured by contact probe coordinate measurement machines to assess the predictive capabilities of spatial statistics through principal component analysis. The validity of the approach has not yet been demonstrated on a large population of mistuned rotors representative of operating fleets, a shortcoming addressed in this work. Furthermore, this work improves the existing predictions by applying principal component methods to sets of airfoil (rotor) measurements, thus effectively capturing blade-to-blade spatial correlations. In conjunction with bootstrap sampling, the method is validated with a set of 40 rotors and quantifies the subset size needed to characterize the population. The work combines a novel statistical representation of rotor geometric mistuning with that of probabilistic techniques to predict the known distribution of forced response amplitudes.

Committee:

Joseph C. Slater, Ph.D., P.E. (Advisor); Jeffrey M. Brown, Ph.D. (Committee Member); J. Mitch Wolff, Ph.D. (Committee Member); Ha-Rok Bae, Ph.D. (Committee Member)

Subjects:

Aerospace Engineering; Engineering; Mechanical Engineering

Keywords:

Principal Component Analysis; PCA; Mistuning; Geometric Mistuning; Probabilistics; Bootstrapping; Turbine Engine; Rotor; Integrally Bladed Rotor; IBR; Blisk; Forced Response; Prediction

Next Page