Validation of Predictive Models: Acceptable Prediction Zone Method
360 likes | 532 Vues
Validation of Predictive Models: Acceptable Prediction Zone Method. Thomas P. Oscar, Ph.D. USDA, Agricultural Research Service Microbial Food Safety Research Unit University of Maryland Eastern Shore Princess Anne, MD. Background Information. Terminology. Performance evaluation
Validation of Predictive Models: Acceptable Prediction Zone Method
E N D
Presentation Transcript
Validation of Predictive Models: Acceptable Prediction Zone Method Thomas P. Oscar, Ph.D. USDA, Agricultural Research Service Microbial Food Safety Research Unit University of Maryland Eastern Shore Princess Anne, MD
Terminology • Performance evaluation • Process of comparing observed and predicted values. • Validation • A potential outcome of performance evaluation. • Requires establishment of criteria.
Test Data Interpolation Extrapolation Performance Bias Accuracy Systematic Bias Criteria
Predictive Modeling Secondary Models Tertiary Model No Model Observed No Predicted No Observed N(t) l Model Observed l Predicted l Primary Model Primary Model mmax Model Observed mmax Predicted mmax Predicted N(t) Predicted N(t) Nmax Model Observed Nmax Predicted Nmax
Performance Evaluation Stage 1 Goodness-of-fit Primary/Secondary Models Verification Tertiary Models Stage 2 Interpolation All Models Stage 3 Extrapolation All Models
Test Data CriteriaInterpolation • Independent data. • Within the response surface. • Uniform coverage. • Collected with same methods. Incomplete and biased evaluation Model data (10 to 40C) versus Test data (25 to 40C)
Test Data CriteriaExtrapolation • Independent data. • Outside the response surface. • Only one variable differs. • Collected with same methods. Confounded comparison Strain A in broth versus Strain B in food
Relative Error (RE) RE for = (predicted - observed)/predicted RE for N(t), No, max and Nmax = (observed - predicted)/predicted RE < 0 are “fail-safe” RE > 0 are “fail-dangerous”
Performance Criteria • Acceptable Predictions -0.30 < RE < 0.15 for mmax -0.60 < RE < 0.30 for l -0.80 < RE < 0.40 for N(t), No, Nmax • Acceptable Performance %RE => 70
Model Development Design • Salmonella Typhimurium • No = 4.8 log CFU/g • Sterile cooked chicken • 10, 12, 14, 16, 20, 24, 28, 32, 36, 38, 40C • Viable counts • BHI agar • 12 per growth curve
Performance Evaluation DesignSecondary Models (Interpolation) • Salmonella Typhimurium • No = 4.8 log CFU/g • Sterile cooked chicken • 11, 13, 15, 18, 22, 26, 30, 34, 37, 39C • Viable counts • BHI agar • 12 per growth curve
Primary ModelLogistic with Delay N = No if t N = Nmax/(1+[(Nmax/No)-1]exp[-max (t-)]) if t >
Secondary Model for No No = mean No
Secondary Model for lHyperbola with Shape Factor = [41.47/(T - 7.325)]1.44
Secondary Model for mmaxModified Square Root max = 0.01885 if T 11.43 max = 0.01885 + [0.004325(T – 11.43)]1.306if T > 11.43
Secondary Model for NmaxAsymptote Model Nmax = exp(2.348[((T – 9.64)(T – 40.74))/((T – 9.606)(T – 40.76))])
Predictive Modeling Secondary Models Tertiary Model No Model Observed No Predicted No Observed N(t) l Model Observed l Predicted l Primary Model Primary Model mmax Model Observed mmax Predicted mmax Predicted N(t) Predicted N(t) Nmax Model Observed Nmax Predicted Nmax
Tertiary Model PerformanceVerification %RE = 90.7
Comparison of Models Fisher’s exact test; P = 0.48, not significant.
Performance Evaluation DesignTertiary Model (Interpolation) • Salmonella Typhimurium • No = 4.8 log CFU/g • Sterile cooked chicken • 11, 13, 15, 18, 22, 26, 30, 34, 37, 39C • Viable counts • BHI agar • 4 per growth curve
Tertiary Model PerformanceInterpolation %RE = 97.5
Should the validated tertiary model be used to predict chicken safety? • Evaluation for extrapolation to: • other initial densities (No) • other strains • other chicken products
Performance Evaluation DesignTertiary Model (Extrapolation) • Salmonella Typhimurium • No = 0.8 log CFU/g • Sterile cooked chicken • 10, 12, 14, 16, 20, 24, 28, 32, 36, 40C • Viable counts • BHI agar • 4 per growth curve
Conclusions • Criteria are important for evaluating performance of models. • Consensus on validation would improve the quality and use of predictive models in the food industry.