Published

2018-07-01

Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression

Usando un anclaje para mejorar predicciones lineales con aplicación a la predicción de progresión de enfermedad

DOI:

https://doi.org/10.15446/rce.v41n2.68535

Keywords:

Anchor, Amyotrophic lateral sclerosis, Biased regression, Linear models, Ordinary least squares (en)
Anclaje, esclerosis lateral amiotrófica, modelos lineales, mínimos cuadrados ordinarios, regresión sesgada (es)

Downloads

Authors

  • Alex G. Karanevich University of Kansas Medical School
  • Jianghua He University of Kansas Medical Center
  • Byron Gajewski University of Kansas Medical Center
Linear models are some of the most straightforward and commonly used modelling approaches. Consider modelling approximately monotonic response data arising from a time-related process. If one has knowledge as to when the process began or ended, then one may be able to leverage additional
assumed data to reduce prediction error. This assumed data, referred to as the anchor, is treated as an additional data-point generated at either the beginning or end of the process. The response value of the anchor is equal to an intelligently selected value of the response (such as the upper bound, lower bound, or 99th percentile of the response, as appropriate). The anchor reduces the variance of prediction at the cost of a possible increase in prediction bias, resulting in a potentially reduced overall mean-square prediction error. This can be extremely eective when few individual data-points are available, allowing one to make linear predictions using as little as a single observed data-point. We develop the mathematics showing the conditions under which an anchor can improve predictions, and also demonstrate using this approach to reduce prediction error when modelling the disease progression of patients with amyotrophic lateral sclerosis.
Modelos lineales son los modelos más fáciles de usar y comunes en modelamiento. Si se considera el modelamiento de una respuesta aprosimadamente monótona que surge de un proceso relacionado al tiempo y se sabe cuándo el proceso inició o terminó, es posible asumir datos adicionales como palanca para reducir el error de predicción. Estos datos adicionales son llamados de anclaje'' y son datos generados antes del inicion o después del final del proceso. El valor de respuesta del anclaje es igual a un valor de respuesta escogido de manera inteligente (como por ejemplo la cota superior, iferior o el percentil 99, según conveniencia). Este anclaje reduce la varianza de la predicción a costo de un posible sesgo en la misma, lo cual resulta en una reducción potencial del error medio de predicción. Lo anterior puede ser extremadamente efectivo cuando haypocos datos individuales, permitiendo hacer predicciones con muy pocos datos. En este trabajo presentamos en desarrollo matemático demostrando las condiciones bajo las cuales el anclaje puede mejorar predicciones y también demostramos una reducción del error de predicción aplicando el método a la modelación de progresión de enfermedad en pacientes con esclerosis lateral amiotrófica.

References

Amemiya, T. (1973), 'Regression analysis when the dependent variable is truncated normal', Econometrica 41(6), 997-1016.

Armon, C., Graves, M., Moses, D., Forté, D., Sepulveda, L., Darby, S. & Smith, R. (2000), 'Linear estimates of disease progression predict survival in patients with amyotrophic lateral sclerosis', Muscle & Nerve 23(6), 874-882.

Atassi, N., Berry, J., Shui, A., Zach, N., Sherman, A., Sinani, E., Walker, J.,

Katsovskiy, I., Schoenfeld, D., Cudkowicz, M. & Leitner, M. (2014), 'The

pro-act database: design, initial analyses, and predictive features', Neurology 83(19), 1719-1725.

Caruana, R., Lou, Y., Gehrke, J., Koch, P., Sturm, M. & Elhadad, N. (2015),

Intelligible models for healthcare: Predicting pneumonia risk and hospital

-day readmission, in 'Proceedings of the 21th ACM SIGKDD International

Conference on Knowledge Discovery and Data Mining', ACM, pp. 1721-1730.

Cedarbaum, J., Stambler, N., Malta, E., Fuller, C., Hilt, D., Thurmond, B. &

Nakanishi, A. (1999), 'The alsfrs-r: a revised als functional rating scale that

incorporates assessments of respiratory function', Journal of the Neurological Sciences 169(1), 13-21.

Gelman, A. (2014), Bayesian data analysis, tercera edn, CRC Press, Boca Raton, FL.

Hoerl, A. & Kennard, R. (2000), 'Ridge regression: Biased estimation for

nonorthogonal problems', Technometrics 42(1), 80-86.

Karanevich, A., Statland, J., Gajewski, B. & He, J. (2018), 'Using an onsetanchored bayesian hierarchical model to improve predictions for amyotrophic lateral sclerosis disease progression', BMC Medical Research Methodology 18(1), 19.

Kutner, M., Nachtsheim, C. & Neter, J. (2004), Applied linear regression models, cuarta edn, McGraw-Hill, New York.

Lesaffre, E., Rizopoulos, D. & Tsonaka, R. (2007), 'The logistic transform for

bounded outcome scores', Biostatistics 8(1), 72-85.

Magnus, T., Beck, M., Giess, R., Puls, I., Naumann, M. & Toyka, K. (2002), 'Disease progression in amyotrophic lateral sclerosis: Predictors of survival',

Muscle & Nerve 25(5), 709-714.

Morris, C. & Lysy, M. (2012), 'Shrinkage estimation in multilevel normal models', Statistical Science 27(1), 115-134.

How to Cite

APA

Karanevich, A. G., He, J. and Gajewski, B. (2018). Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression. Revista Colombiana de Estadística, 41(2), 137–155. https://doi.org/10.15446/rce.v41n2.68535

ACM

[1]
Karanevich, A.G., He, J. and Gajewski, B. 2018. Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression. Revista Colombiana de Estadística. 41, 2 (Jul. 2018), 137–155. DOI:https://doi.org/10.15446/rce.v41n2.68535.

ACS

(1)
Karanevich, A. G.; He, J.; Gajewski, B. Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression. Rev. colomb. estad. 2018, 41, 137-155.

ABNT

KARANEVICH, A. G.; HE, J.; GAJEWSKI, B. Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression. Revista Colombiana de Estadística, [S. l.], v. 41, n. 2, p. 137–155, 2018. DOI: 10.15446/rce.v41n2.68535. Disponível em: https://revistas.unal.edu.co/index.php/estad/article/view/68535. Acesso em: 28 mar. 2025.

Chicago

Karanevich, Alex G., Jianghua He, and Byron Gajewski. 2018. “Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression”. Revista Colombiana De Estadística 41 (2):137-55. https://doi.org/10.15446/rce.v41n2.68535.

Harvard

Karanevich, A. G., He, J. and Gajewski, B. (2018) “Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression”, Revista Colombiana de Estadística, 41(2), pp. 137–155. doi: 10.15446/rce.v41n2.68535.

IEEE

[1]
A. G. Karanevich, J. He, and B. Gajewski, “Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression”, Rev. colomb. estad., vol. 41, no. 2, pp. 137–155, Jul. 2018.

MLA

Karanevich, A. G., J. He, and B. Gajewski. “Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression”. Revista Colombiana de Estadística, vol. 41, no. 2, July 2018, pp. 137-55, doi:10.15446/rce.v41n2.68535.

Turabian

Karanevich, Alex G., Jianghua He, and Byron Gajewski. “Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression”. Revista Colombiana de Estadística 41, no. 2 (July 1, 2018): 137–155. Accessed March 28, 2025. https://revistas.unal.edu.co/index.php/estad/article/view/68535.

Vancouver

1.
Karanevich AG, He J, Gajewski B. Using an Anchor to Improve Linear Predictions with Application to Predicting Disease Progression. Rev. colomb. estad. [Internet]. 2018 Jul. 1 [cited 2025 Mar. 28];41(2):137-55. Available from: https://revistas.unal.edu.co/index.php/estad/article/view/68535

Download Citation

CrossRef Cited-by

CrossRef citations1

1. Divya Ramamoorthy, Kristen Severson, Soumya Ghosh, Karen Sachs, Emily G. Baxi, Alyssa N. Coyne, Elizabeth Mosmiller, Lindsey Hayes, Aianna Cerezo, Omar Ahmad, Promit Roy, Steven Zeiler, John W. Krakauer, Jonathan Li, Aneesh Donde, Nhan Huynh, Miriam Adam, Brook T. Wassie, Alex Lenail, Natasha Leanna Patel-Murray, Yogindra Raghav, Karen Sachs, Velina Kozareva, Stanislav Tsitkov, Tobias Ehrenberger, Julia A. Kaye, Leandro Lima, Stacia Wyman, Edward Vertudes, Naufa Amirani, Krishna Raja, Reuben Thomas, Ryan G. Lim, Ricardo Miramontes, Jie Wu, Vineet Vaibhav, Andrea Matlock, Vidya Venkatraman, Ronald Holewenski, Niveda Sundararaman, Rakhi Pandey, Danica-Mae Manalo, Aaron Frank, Loren Ornelas, Lindsey Panther, Emilda Gomez, Erick Galvez, Daniel Perez, Imara Meepe, Susan Lei, Louis Pinedo, Chunyan Liu, Ruby Moran, Dhruv Sareen, Barry Landin, Carla Agurto, Guillermo Cecchi, Raquel Norel, Sara Thrower, Sarah Luppino, Alanna Farrar, Lindsay Pothier, Hong Yu, Ervin Sinani, Prasha Vigneswaran, Alexander V. Sherman, S. Michelle Farr, Berhan Mandefro, Hannah Trost, Maria G. Banuelos, Veronica Garcia, Michael Workman, Richie Ho, Robert Baloh, Jennifer Roggenbuck, Matthew B. Harms, Carolyn Prina, Sarah Heintzman, Stephen Kolb, Jennifer Stocksdale, Keona Wang, Todd Morgan, Daragh Heitzman, Arish Jamil, Jennifer Jockel-Balsarotti, Elizabeth Karanja, Jesse Markway, Molly McCallum, Tim Miller, Ben Joslin, Deniz Alibazoglu, Senda Ajroud-Driss, Jay C. Beavers, Mary Bellard, Elizabeth Bruce, Nicholas Maragakis, Merit E. Cudkowicz, James Berry, Terri Thompson, Steven Finkbeiner, Leslie M. Thompson, Jennifer E. Van Eyk, Clive N. Svendsen, Jeffrey D. Rothstein, Jonathan D. Glass, Christina N. Fournier, Alexander Sherman, Christian Lunetta, David Walk, Ghazala Hayat, James Wymer, Kelly Gwathmey, Nicholas Olney, Senda Ajroud-Driss, Terry Heiman-Patterson, Ximena Arcila-Londono, Kenneth Faulconer, Ervin Sanani, Alex Berger, Julia Mirochnick, Todd M. Herrington, James D. Berry, Kenney Ng, Ernest Fraenkel. (2022). Identifying patterns in amyotrophic lateral sclerosis progression from sparse longitudinal data. Nature Computational Science, 2(9), p.605. https://doi.org/10.1038/s43588-022-00299-w.

Dimensions

PlumX

  • Citations
  • CrossRef - Citation Indexes: 2
  • Scopus - Citation Indexes: 1
  • Usage
  • SciELO - Full Text Views: 261
  • SciELO - Abstract Views: 31
  • Captures
  • Mendeley - Readers: 2
  • Social Media
  • Facebook - Shares, Likes & Comments: 8

Article abstract page views

695

Downloads