Douglas M. Hawkins
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Douglas M. Hawkins has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
- David J. Olive
Department of Mathematics, Southern Illinois University, Mailcode 4408, Carbondale, IL 62901-4408, USA
2 shared publications Denise Mills
Natural Resources Research Institute, University of Minnesota Duluth, 5013 Miller Trunk Highway, Duluth, MN 55811, USA
2 shared publications- Andrew K. Godwin
Department of Pathology and Laboratory Medicine, University of Kansas Medical Center, Kansas City, USA
1 shared publication Betty Stephenson
Becton Dickinson Bioscience, Sparks, MD 21152‐0999, U.S.A.
1 shared publication- Binbin Yue
Program in Oncology, University of Maryland Greenebaum Comprehensive Cancer Center, Baltimore, MD 21201, USA
1 shared publication - Chang Wook Kang
Hanyang University, Seoul, Korea
1 shared publication
Latest Publications
- Open Access
- 8 February 2012
- Published by Springer Nature in Breast Cancer Research
- Vol. 14 (1) , R26
Progranulin (GP88) tumor tissue expression is associated with increased risk of recurrence in breast cancer patients diagnosed with estrogen receptor positive invasive ductal carcinoma
Introduction: GP88 (progranulin) has been implicated in tumorigenesis and resistance to anti-estrogen therapies for estrogen receptor positive (ER+) breast cancer. Previous pathological studies showed that GP88 is expressed in invasive ductal carcinoma (IDC), but not in normal mammary epithelial tissue, benign lesions or lobular carcinoma. Based on these results, the present study examines GP88 prognostic significance in association with recurrence and death risks for ER+ IDC patients. Methods: Two retrospective multi-site clinical studies examined GP88 expression by immunohistochemistry (IHC) analysis of paraffin-embedded breast tumor tissue sections from ER+ IDC patients (lymph node positive and negative, stage 1 to 3) in correlation with patients' survival outcomes. The training study established a GP88 cut-off value associated with decreased disease-free (DFS) and overall (OS) survivals. The validation study verified the GP88 cut-off value and compared GP88 prognostic information with other prognostic factors, particularly tumor size, grade, disease stage and lymph node status in multivariate analysis. Results: GP88 expression is associated with a statistically significant increase in recurrence risk for ER+ IDC patients. The training study established that GP88 3+ score was associated with decreased DFS (P = 0.0004) and OS (P = 0.0036). The independent validation study verified that GP88 3+ score was associated with a 5.9-fold higher hazard of disease recurrence and a 2.5-fold higher mortality hazard compared to patients with tumor GP88 < 3+. GP88 remained an independent risk predictor after considering age, ethnicity, nodal status, tumor size, tumor grade, disease stage, progesterone receptor expression and treatments. Conclusions: The survival factor GP88 is a novel prognostic biomarker, predictive of recurrence risk and increased mortality for non-metastatic ER+ IDC patients. Of importance, our data show that GP88 continues to be a prognostic factor even after five years. These results also provide evidence that GP88 provides prognostic information independent of tumor and clinical characteristics and would support prospective study to examine whether GP88 expression could help stratify patients with ER+ tumors for adjuvant therapy.
- 8 February 2012
- Published by Springer Nature in Breast Cancer Research
- Vol. 14 (1) , R26
Open AccessProgranulin (GP88) tumor tissue expression is associated with increased risk of recurrence in breast cancer patients diagnosed with estrogen receptor positive invasive ductal carcinoma
Introduction: GP88 (progranulin) has been implicated in tumorigenesis and resistance to anti-estrogen therapies for estrogen receptor positive (ER+) breast cancer. Previous pathological studies showed that GP88 is expressed in invasive ductal carcinoma (IDC), but not in normal mammary epithelial tissue, benign lesions or lobular carcinoma. Based on these results, the present study examines GP88 prognostic significance in association with recurrence and death risks for ER+ IDC patients. Methods: Two retrospective multi-site clinical studies examined GP88 expression by immunohistochemistry (IHC) analysis of paraffin-embedded breast tumor tissue sections from ER+ IDC patients (lymph node positive and negative, stage 1 to 3) in correlation with patients' survival outcomes. The training study established a GP88 cut-off value associated with decreased disease-free (DFS) and overall (OS) survivals. The validation study verified the GP88 cut-off value and compared GP88 prognostic information with other prognostic factors, particularly tumor size, grade, disease stage and lymph node status in multivariate analysis. Results: GP88 expression is associated with a statistically significant increase in recurrence risk for ER+ IDC patients. The training study established that GP88 3+ score was associated with decreased DFS (P = 0.0004) and OS (P = 0.0036). The independent validation study verified that GP88 3+ score was associated with a 5.9-fold higher hazard of disease recurrence and a 2.5-fold higher mortality hazard compared to patients with tumor GP88 < 3+. GP88 remained an independent risk predictor after considering age, ethnicity, nodal status, tumor size, tumor grade, disease stage, progesterone receptor expression and treatments. Conclusions: The survival factor GP88 is a novel prognostic biomarker, predictive of recurrence risk and increased mortality for non-metastatic ER+ IDC patients. Of importance, our data show that GP88 continues to be a prognostic factor even after five years. These results also provide evidence that GP88 provides prognostic information independent of tumor and clinical characteristics and would support prospective study to examine whether GP88 expression could help stratify patients with ER+ tumors for adjuvant therapy.
- 1 March 2010
- Published by Wiley in Journal of Chemometrics
- Vol. 24 (3-4) , 188-193
Deterministic fallacies and model validation
Stochastic settings differ from deterministic ones in many subtle ways, making it easy to slip into errors through applying deterministic thinking inappropriately. We suspect this is the cause of much of the disagreement about model validation. A further technical issue is a common misapplication of cross‐validation, in which it is applied only partially, leading to incorrect results. Statistical theory and empirical investigation verify the efficacy of cross‐validation when it is applied correctly. In settings where data are relatively scarce, cross‐validation is attractive in that it makes the maximum possible use of all available information, at the cost of potentially substantial computation. The bootstrap is another method that makes full use of all available data for both model fitting and model validation, at a cost of substantially increased computation, and it shares many of the broad philosophical background of cross‐validation. Increasingly, the computational cost of these methods is not a major concern, leading to the recommendation, in most circumstances, to use cross‐validation or bootstrapping rather than the earlier standard method of splitting the available data into a learning and a testing portion. Copyright © 2010 John Wiley & Sons, Ltd.
- 1 March 2010
- Published by Wiley in Journal of Chemometrics
- Vol. 24 (3-4) , 188-193
Deterministic fallacies and model validation
Stochastic settings differ from deterministic ones in many subtle ways, making it easy to slip into errors through applying deterministic thinking inappropriately. We suspect this is the cause of much of the disagreement about model validation. A further technical issue is a common misapplication of cross‐validation, in which it is applied only partially, leading to incorrect results. Statistical theory and empirical investigation verify the efficacy of cross‐validation when it is applied correctly. In settings where data are relatively scarce, cross‐validation is attractive in that it makes the maximum possible use of all available information, at the cost of potentially substantial computation. The bootstrap is another method that makes full use of all available data for both model fitting and model validation, at a cost of substantially increased computation, and it shares many of the broad philosophical background of cross‐validation. Increasingly, the computational cost of these methods is not a major concern, leading to the recommendation, in most circumstances, to use cross‐validation or bootstrapping rather than the earlier standard method of splitting the available data into a learning and a testing portion. Copyright © 2010 John Wiley & Sons, Ltd.
- 1 July 2009
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 41 (3) , 285-303
A Multivariate Change-Point Model for Change in Mean Vector and/or Covariance Structure
The motivation of the current research manuscript is to provide practitioners with a multivariate-analysis tool able to detect change in the mean vector and/or covariance matrix, as well as the epoch of a change, in an independent sequence of multivariate observations. The article explores the multivariate change-point model through generalized likelihood-ratio statistics applied sequentially and adapted to repeated use. We sought an analytical result for the exact moments of the generalized likelihood ratio (GLR) statistic. The benefit flowing from this sequential adaptation is to be able to monitor short runs and unknown parameter processes while controlling their run behavior. Possible areas of application are short run processes, sequential dynamic control, ambulatory monitoring, disease monitoring, and syndromic surveillance.
- 1 July 2009
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 41 (3) , 285-303
A Multivariate Change-Point Model for Change in Mean Vector and/or Covariance Structure
The motivation of the current research manuscript is to provide practitioners with a multivariate-analysis tool able to detect change in the mean vector and/or covariance matrix, as well as the epoch of a change, in an independent sequence of multivariate observations. The article explores the multivariate change-point model through generalized likelihood-ratio statistics applied sequentially and adapted to repeated use. We sought an analytical result for the exact moments of the generalized likelihood ratio (GLR) statistic. The benefit flowing from this sequential adaptation is to be able to monitor short runs and unknown parameter processes while controlling their run behavior. Possible areas of application are short run processes, sequential dynamic control, ambulatory monitoring, disease monitoring, and syndromic surveillance.
- 1 March 2007
- Published by Elsevier in Statistics & Probability Letters
- Vol. 77 (6) , 621-624
Behavior of elemental sets in regression
- 1 March 2007
- Published by Elsevier in Statistics & Probability Letters
- Vol. 77 (6) , 621-624
Behavior of elemental sets in regression
- 10 January 2006
- Published by Springer Nature in Metrika
- Vol. 63 (3) , 309-315
On the Inconsistency of the Change-Point Estimator for the NE Family
- 10 January 2006
- Published by Springer Nature in Metrika
- Vol. 63 (3) , 309-315
On the Inconsistency of the Change-Point Estimator for the NE Family
- 31 March 2004
- Published by Elsevier in Environmental Toxicology and Pharmacology
- Vol. 16 (1-2) , 37-44
QSARs for chemical mutagens from structure: ridge regression fitting and diagnostics
- 31 March 2004
- Published by Elsevier in Environmental Toxicology and Pharmacology
- Vol. 16 (1-2) , 37-44
QSARs for chemical mutagens from structure: ridge regression fitting and diagnostics
- Top Cited Papers
- 2 December 2003
- Published by American Chemical Society (ACS) in Journal of Chemical Information and Computer Sciences
- Vol. 44 (1) , 1-12
The Problem of Overfitting
- 2 December 2003
- Published by American Chemical Society (ACS) in Journal of Chemical Information and Computer Sciences
- Vol. 44 (1) , 1-12
Top Cited PapersThe Problem of Overfitting
- 1 October 2003
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 35 (4) , 355-366
The Changepoint Model for Statistical Process Control
Statistical process control (SPC) requires statistical methodologies that detect changes in the pattern of data over time. The common methodologies, such as Shewhart, cumulative sum (cusum), and exponentially weighted moving average (EWMA) charting, require the in-control values of the process parameters, but these are rarely known accurately. Using estimated parameters, the run length behavior changes randomly from one realization to another, making it impossible to control the run length behavior of any particular chart. A suitable methodology for detecting and diagnosing step changes based on imperfect process knowledge is the unknown-parameter changepoint formulation. Long recognized as a Phase I analysis tool, we argue that it is also highly effective in allowing the user to progress seamlessly from the start of Phase I data gathering through Phase II SPC monitoring. Despite not requiring specification of the post-change process parameter values, its performance is never far short of that of the optimal cusum chart which requires this knowledge, and it is far superior for shifts away from the cusum shift for which the cusum chart is optimal. As another benefit, while changepoint methods are designed for step changes that persist, they are also competitive with the Shewhart chart, the chart of choice for isolated non-sustained special causes.
- 1 October 2003
- Published by Taylor & Francis in Journal of Quality Technology
- Vol. 35 (4) , 355-366
The Changepoint Model for Statistical Process Control
Statistical process control (SPC) requires statistical methodologies that detect changes in the pattern of data over time. The common methodologies, such as Shewhart, cumulative sum (cusum), and exponentially weighted moving average (EWMA) charting, require the in-control values of the process parameters, but these are rarely known accurately. Using estimated parameters, the run length behavior changes randomly from one realization to another, making it impossible to control the run length behavior of any particular chart. A suitable methodology for detecting and diagnosing step changes based on imperfect process knowledge is the unknown-parameter changepoint formulation. Long recognized as a Phase I analysis tool, we argue that it is also highly effective in allowing the user to progress seamlessly from the start of Phase I data gathering through Phase II SPC monitoring. Despite not requiring specification of the post-change process parameter values, its performance is never far short of that of the optimal cusum chart which requires this knowledge, and it is far superior for shifts away from the cusum shift for which the cusum chart is optimal. As another benefit, while changepoint methods are designed for step changes that persist, they are also competitive with the Shewhart chart, the chart of choice for isolated non-sustained special causes.
- 1 July 2003
- Published by Elsevier in Statistics & Probability Letters
- Vol. 63 (3) , 259-266
Robust regression with high coverage
- 1 July 2003
- Published by Elsevier in Statistics & Probability Letters
- Vol. 63 (3) , 259-266
Robust regression with high coverage
- Top Cited Papers
- 24 January 2003
- Published by American Chemical Society (ACS) in Journal of Chemical Information and Computer Sciences
- Vol. 43 (2) , 579-586
Assessing Model Fit by Cross-Validation
When QSAR models are fitted, it is important to validate any fitted modelto check that it is plausible that its predictions will carry over to fresh data not used in the model fitting exercise. There are two standard ways of doing thisusing a separate hold-out test sample and the computationally much more burdensome leave-one-out cross-validation in which the entire pool of available compounds is used both to fit the model and to assess its validity. We show by theoretical argument and empiric study of a large QSAR data set that when the available sample size is smallin the dozens or scores rather than the hundreds, holding a portion of it back for testing is wasteful, and that it is much better to use cross-validation, but ensure that this is done properly.
- 24 January 2003
- Published by American Chemical Society (ACS) in Journal of Chemical Information and Computer Sciences
- Vol. 43 (2) , 579-586
Top Cited PapersAssessing Model Fit by Cross-Validation
When QSAR models are fitted, it is important to validate any fitted modelto check that it is plausible that its predictions will carry over to fresh data not used in the model fitting exercise. There are two standard ways of doing thisusing a separate hold-out test sample and the computationally much more burdensome leave-one-out cross-validation in which the entire pool of available compounds is used both to fit the model and to assess its validity. We show by theoretical argument and empiric study of a large QSAR data set that when the available sample size is smallin the dozens or scores rather than the hundreds, holding a portion of it back for testing is wasteful, and that it is much better to use cross-validation, but ensure that this is done properly.