Predicting response to pembrolizumab in metastatic melanoma by a new personalization algorithm

Table 8 Confusion table for the classification of the model parameter, \(\gamma_{mel}\)

\(\kappa = 0.222\)	Fitted value of \(\gamma_{mel}\)
\(\kappa = 0.222\)	Value 1	Value 2
Classification-derived value of \(\gamma_{mel}\)
Value 1	19	13
Value 2	8	14

The parameter \(\gamma_{mel}\) can take on two numerical values, namely 0.005 day⁻¹ (value 1), or 0.003269 day⁻¹ (value 2). Presented in the table are the number of cases in which value 1 of \(\gamma_{mel}\) was fitted, and identified as value 1 by the classification algorithm (true positive; upper-left cell), or value 2 (false negative; lower left cell), and the number of cases in which value 2 of \(\gamma_{mel}\) was fitted, as well as classified as value 2 (true negative; right lower cell), or misclassified as value 1 (false positive; right upper cell). The calculated Cohen’s kappa, \(\kappa ,\) suggests a fair agreement of predictions to fitting results

ISSN: 1479-5876