Predictive model for the growth rate of tomatoes in saline substrate cultivation

Soil salinization which has a detrimental effect on the fruit/crop productivity over time, is exacerbated by the climate change phenomena. To counteract the salinity effect, the HaloFarMs project experiments with tomatoes cultivated in monoculture and intercropped with halophytes. Based on that experimental activity, this study uses tools from data mining to build a joint probability mass function of the initial salinity of the soil and the growth rate of the fruits. Their mutual dependence is quantified with the help of tools from information theory. It turns out that their dependence is significant. Based on this discovery, this study uses non-linear regression analysis to match the first moments of the joint probability mass function which are the average electrical conductivity and the average growth rate (either number of mature fruits or their weight change per day). The resulting sigmoid function has a coefficient of determination of more than 99 per cent.