Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A Personalized Predictive Model That Jointly Optimizes Discrimination and Calibration.

Tatiana Krikella1, Joel A Dubin1

  • 1Department of Statistics and Actuarial Science, University of Waterloo, Waterloo, Ontario, Canada.

Statistics in Medicine
|May 16, 2025
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Use of and Barriers to Mental Health Services for Youth with Physical-Mental Multimorbidity: A Nationally Representative Study.

The journal of behavioral health services & research·2026
Same author

COVID-19 testing and mental health service utilization in Ontario: a population-based cohort study.

International journal of public health·2026
Same author

Health Functioning Moderates the Association Between Chronic Conditions and Mental Illness in Childhood.

Child: care, health and development·2026
Same author

Unsupervised identification of sepsis subpopulations in the eICU database: A multi-method clustering approach with validation.

Computers in biology and medicine·2026
Same author

Youth stress, happiness, and life satisfaction across morbidity status: a gender-stratified analysis.

Scientific reports·2026
Same author

The Impact of Calorie Menu Labeling on Disordered Eating and Related Psychosocial Outcomes: A Longitudinal Study Among Young Adults in Canada.

American journal of preventive medicine·2026
Same journal

A Mixture of Distributed Lag Non-Linear Models to Account for Spatially Heterogeneous Exposure-Lag-Response Associations.

Statistics in medicine·2026
Same journal

Practical Considerations for Gaussian Process Modeling for Causal Inference in Quasi-Experimental Studies With Panel Data.

Statistics in medicine·2026
Same journal

Covariate Adjustment for Wilcoxon Two Sample Statistic and Test.

Statistics in medicine·2026
Same journal

Beyond Fixed Thresholds: Optimizing Summaries of Wearable Device Data via Piecewise Linearization of Quantile Functions.

Statistics in medicine·2026
Same journal

A Causal Framework for Evaluating the Total Effect of Strategies Aiming to Expand Screening and to Improve Outcomes.

Statistics in medicine·2026
Same journal

Causal Effects on Nonterminal Event Time With Application to Antibiotic Usage and Future Resistance.

Statistics in medicine·2026
See all related articles

This study introduces a new algorithm for personalized predictive models (PPMs) in precision medicine. It optimizes both model discrimination and calibration by selecting an ideal similar subpopulation size, improving patient health predictions.

Area of Science:

  • Health Informatics
  • Biostatistics
  • Computational Biology

Background:

  • Precision medicine leverages patient similarity for improved predictive modeling.
  • Current models often prioritize discrimination over calibration, potentially leading to misleading results.
  • Assessing model calibration is crucial but frequently overlooked in health research.

Purpose of the Study:

  • To propose an algorithm for fitting personalized predictive models (PPMs) using an optimal similar subpopulation size.
  • To jointly optimize model discrimination and calibration, addressing the limitations of current approaches.
  • To introduce a flexible mixture loss function for balancing discrimination and calibration.

Main Methods:

  • Developed a novel algorithm to determine the optimal size of a similar subpopulation for PPMs.
Keywords:
Brier Scorecosine similaritymixture loss functionprecision medicineprediction modelsubpopulation

Related Experiment Videos

  • Defined a mixture loss function incorporating both discrimination and calibration metrics.
  • Empirically investigated the quadratic relationship between subpopulation size and model calibration.
  • Analyzed the impact of within-population patient weighting on predictive performance.

    • Main Results:

    • The proposed algorithm effectively optimizes both discrimination and calibration in PPMs.
    • A quadratic relationship was identified between subpopulation size and model calibration.
    • Subpopulation size demonstrated a greater impact on PPM performance than patient weighting functions.
    • The mixture loss function allows for tunable emphasis on discrimination versus calibration.

    Conclusions:

    • The developed algorithm enhances the reliability of personalized predictive models in precision medicine.
    • Jointly optimizing discrimination and calibration leads to more trustworthy and accurate patient predictions.
    • Subpopulation size is a critical factor in achieving well-calibrated and discriminative PPMs.
    • Future research should consider the interplay between subpopulation selection and model performance metrics.