Physical Sciences and Mathematics Commons^™

The Validity Of Online Patient Ratings Of Physicians, Jennifer L. Priestley, Yiyun Zhou, Robert Mcgrath

Jennifer L. Priestley

Background: Information from ratings sites are increasingly informing patient decisions related to health care and the selection of physicians.

Objective: The current study sought to determine the validity of online patient ratings of physicians through comparison with physician peer review.

Methods: We extracted 223,715 reviews of 41,104 physicians from 10 of the largest cities in the United States, including 1142 physicians listed as “America’s Top Doctors” through physician peer review. Differences in mean online patient ratings were tested for physicians who were listed and those who were not.

Results: Overall, no differences were found between the online patient ratings based …

Logistic Ensemble Models, Bob Vanderheyden, Jennifer L. Priestley

Jennifer L. Priestley

Predictive models that are developed in a regulated industry or a regulated application, like determination of credit worthiness must be interpretable and “rational” (e.g., improvements in basic credit behavior must result in improved credit worthiness scores). Machine Learning technologies provide very good performance with minimal analyst intervention, so they are well suited to a high volume analytic environment but the majority are “black box” tools that provide very limited insight or interpretability into key drivers of model performance or predicted model output values. This paper presents a methodology that blends one of the most popular predictive statistical modeling methods with …

Influence Of The Event Rate On Discrimination Abilities Of Bankruptcy Prediction Models, Lili Zhang, Jennifer Priestley, Xuelei Ni

Jennifer L. Priestley

In bankruptcy prediction, the proportion of events is very low, which is often oversampled to eliminate this bias. In this paper, we study the influence of the event rate on discrimination abilities of bankruptcy prediction models. First the statistical association and significance of public records and firmographics indicators with the bankruptcy were explored. Then the event rate was oversampled from 0.12% to 10%, 20%, 30%, 40%, and 50%, respectively. Seven models were developed, including Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, Support Vector Machine, Bayesian Network, and Neural Network. Under different event rates, models were comprehensively evaluated and compared …

Application Of Isotonic Regression In Predicting Business Risk Scores, Linh T. Le, Jennifer L. Priestley

Jennifer L. Priestley

An isotonic regression model fits an isotonic function of the explanatory variables to estimate the expectation of the response variable. In other words, as the function increases, the estimated expectation of the response must be non-decreasing. With this characteristic, isotonic regression could be a suitable option to analyze and predict business risk scores. A current challenge of isotonic regression is the decrease of performance when the model is fitted in a large data set e.g. more than four or five dimensions. This paper attempts to apply isotonic regression models into prediction of business risk scores using a large data set …

A Comparison Of Decision Tree With Logistic Regression Model For Prediction Of Worst Non-Financial Payment Status In Commercial Credit, Jessica M. Rudd Mph, Gstat, Jennifer L. Priestley

Jennifer L. Priestley

Credit risk prediction is an important problem in the financial services domain. While machine learning techniques such as Support Vector Machines and Neural Networks have been used for improved predictive modeling, the outcomes of such models are not readily explainable and, therefore, difficult to apply within financial regulations. In contrast, Decision Trees are easy to explain, and provide an easy to interpret visualization of model decisions. The aim of this paper is to predict worst non-financial payment status among businesses, and evaluate decision tree model performance against traditional Logistic Regression model for this task. The dataset for analysis is provided …

Binary Classification On Past Due Of Service Accounts Using Logistic Regression And Decision Tree, Yan Wang, Jennifer L. Priestley

Jennifer L. Priestley

This paper aims at predicting businesses’ past due in service accounts as well as determining the variables that impact the likelihood of repayment. Two binary classification approaches, logistic regression and the decision tree, were conducted and compared. Both approaches have very good performances with respect to the accuracy. However, the decision tree only uses 10 predictors and reaches an accuracy of 96.69% on the validation set while logistic regression includes 14 predictors and reaches an accuracy of 94.58%. Due to the large concern of false negatives in financial industry, the decision tree technique is a better option than logistic regression …

A Comparison Of Machine Learning Algorithms For Prediction Of Past Due Service In Commercial Credit, Liyuan Liu M.A, M.S., Jennifer Lewis Priestley Ph.D.

Jennifer L. Priestley

Credit risk modeling has carried a variety of research interest in previous literature, and recent studies have shown that machine learning methods achieved better performance than conventional statistical ones. This study applies decision tree which is a robust advanced credit risk model to predict the commercial non-financial past-due problem with better critical power and accuracy. In addition, we examine the performance with logistic regression analysis, decision trees, and neural networks. The experimenting results confirm that decision trees improve upon other methods. Also, we find some interesting factors that impact the commercials’ non-financial past-due payment.

A Comparison Of Machine Learning Techniques And Logistic Regression Method For The Prediction Of Past-Due Amount, Jie Hao, Jennifer L. Priestley

Jennifer L. Priestley

The aim of this paper to predict a past-due amount using traditional and machine learning techniques: Logistic Analysis, k-Nearest Neighbor and Random Forest. The dataset to be analyzed is provided by Equifax, which contains 305 categories of financial information from more than 11,787,287 unique businesses from 2006 to 2014. The big challenge is how to handle with the big and noisy real world datasets. Among the three techniques, the results show that Logistic Regression Method is the best in terms of predictive accuracy and type I errors.

An Analysis Of Accuracy Using Logistic Regression And Time Series, Edwin Baidoo, Jennifer L. Priestley

Jennifer L. Priestley

This paper analyzes the accuracy rates for logistic regression and time series models. It also examines a relatively new performance index that takes into consideration the business assumptions of credit markets. Although prior research has focused on evaluation metrics, such as AUC and Gini index, this new measure has a more intuitive interpretation for various managers and decision makers and can be applied to both Logistic and Time Series models.

Counting The Impossible: Sampling And Modeling To Achieve A Large State Homeless Count, Jennifer L. Priestley, Jane Massey

Jennifer L. Priestley

Objective: Using inferential statistics, we develop estimates of the homeless population of a geographically large and economically diverse state -- Georgia.

Methods: Multiple independent data sources (2000 U.S. Census, the 2006 Georgia County Guide, Georgia Chamber of Commerce) were used to develop Clusters of the 150 Georgia Counties. These clusters were used as "strata" to then execute traified sampling. Homeless counts were conducted within the sample counties, allowing for multiple regression models to be developed to generate predictions of homeless persons by county.

Results: In response to a mandate from the US Department of Housing and Urban Development, the State …

Multi-Organizational Networks: Three Antecedents Of Knowledge Transfer, Jennifer L. Priestley, Subhashish Samaddar

Jennifer L. Priestley

Researchers have demonstrated that organizations operating within formal networks are more likely to experience knowledge transfer, and the associated benefits of knowledge transfer, than would organizations operating outside of a network. However, limited research attention has been given to how the established antecedents of knowledge transfer are affected by the different forms that multi-organizational networks can assume. Using two case studies, we develop six testable propositions regarding how three of the established antecedents of knowledge transfer —absorptive capacity, shared identity and causal ambiguity—would be affected by the different characteristics, which define multi-organizational network form. We discuss these propositions and raise …