Adjusted Prediction of Plant Height with Application to Greenhouse Grown Poinsettias

The objective is to predict future plant growth using data from greenhouse grown poinsettias. A population‐mean growth trajectory can be predicted from growth conditions using previously published results. Four different predictors are used to identify the local group‐effects that come in addition to the population‐mean response to growth conditions, in order to improve the predictions of the final plant height for specific plant groups. A search for optimal model complexity and use of growth history during calibration is conducted using cross‐validation.