We implement the out-of-sample prediction from the first quarter of 2010 to the fourth quarter of 2015.
These measures were either based on predictions from a model fit on the full dataset or, as a sensitivity analysis, on out-of-sample predictions from leave-one-hospital-out cross-validation as described above.
Specifically, net reclassification improvement and integrated discrimination improvement were 0.17 (P < 0.001) and 0.04 (P < 0.001), respectively, if based on predictions derived on the full dataset, and 0.13 (P = 0.01) and 0.04 (P < 0.001), if based on out-of-sample predictions from leave-one-hospital out cross-validation.
Using this model, and including the same economic and other factors as in the original model, I can again generate an out-of-sample prediction for each president from Truman through Mr. Obama (click image to enlarge): This model's predictions are not consistently greater than actual approval.
Fig. 15 Sample prediction results (unpredictable pattern).
By using early versus late stage disease for out of sample predictions, the signature derived from early stage cancers could accurately (86%) predict mutation status of late stage cancers.
These studies have been accomplished through shot-gun sequencing of cDNAs obtained from size-fractionated RNA samples, computational prediction from genomic data, or a combination of both [ 26].
Inferences concerning future longitudinal values from a new individual (an out-of-sample prediction) can be obtained from the posterior predictive distribution.
As a final test, we performed two out-of-sample predictions using our estimates from the 1995 2007 period.
Normal (N) and tumor (T) sample sizes, accuracy of prediction from binary tree algorithm (% accuracy), and the number of genes in classifier (number of genes) were shown for each study, separately.
