This research investigates and compares artificial neural network and multiple linear regression for predicting the chlorophyll concentration index of Saint John’s wort leaves (Hypericum perforatum L.). Plants were fertilized with 0, 30, 60, 90, and 120 kg ha⁻¹ nitrogen [34% nitrogen ammonium nitrate (NH₄NO₃)]. Chlorophyll concentration index of each leaf was measured using SPAD meter. Afterwards, rgb (red, green, and blue color) values of all leaf images were determined by image processing. Values obtained were modeled using both multiple regression analysis and artificial neural networks. Using multiple regression analysis R² values were between 0.61 and 0.97. Coefficient of determination values (R²) using artificial neutral network values were found to be 0.99. Artificial neutral network modeling successfully described the relationship between actual chlorophyll concentration index values and predicted chlorophyll concentration index values.

The quantitative effects of temperature and light intensity on accumulation of phenolics were examined on greenhouse-grown plants of Hypericum perforatum L. Plants were grown in a greenhouse separated into two parts: shaded by 50% transparent polyethylene cover and un-shaded. Temperature values and light intensities were measured daily during the experiment, while plants were harvested weekly for HPLC analyses. Multi regression analyses were performed to describe the quantitative effects of temperature and light intensity on phenolics accumulation. According to the results, increases in temperatures from 24°C to 32°C and light intensities from 803.4 μMm⁻²s⁻¹ to 1618.6 μMm⁻²s⁻¹ resulted in a continuous increase in amentoflavone, apigenin-7-glucoside, cholorogenic acid, hyperoside, kaempferol, rutin, quercetin and quercitrin contents. The relationships between temperature, light intensity and phenolics accumulation were formulized as P= [a + (b₁ x t) + (b₂ x l) + [b₃ x(t x l)]] equition, where P is the content of the corresponding phenolic, t temperature (°C), l light intensity (μMm⁻²s⁻¹) and a, b₁, b₂ and b3 the coefficients of the produced equation. The regression coefficient (R²) value for amentoflavone was 0.84, for apigenin-7-glucoside 0.87, for cholorogenic acid 0.83, for hyperoside 0.95, for kaempferol 0.76, for rutin 0.70, for quercetin – 0.93, and for quercitrin – 0.86. All R² values and standard errors of the equations were found to be significant at the p<0.001 level. The mathematical models produced in the present study could be applied by Hypericum researchers as useful tools for the prediction of phenolics content instead of routine chemical analyses.

The artificial neural network (ANN) method was used in this study for predicting sugar beet (Beta vulgaris L.) leaf chlorophyll concentration from leaves. The experiment was carried out in field conditions in 2015-2016. In this research, symbiotic mychorrhizae as Bio-one (Azotobacter vinelandii and Clostridium pasteurianum) in commercial preparation (10 kg/da) and ammonium sulfate (40 kg/da) were use used as a fertilizer. In order to measure the leaves' chlorophyll concentration we used a SPAD-502 chlorophyll meter. Artificial neural network, red, green, and blue components of the images were used which was developed to predict chlorophyll concentration. The results showed the ANN model able to estimate sugar beet leaf chlorophyll concentration. The coefficient of determination (R-2) was found to be 0.98 while mean square error (MSE) was obtained as 0.007 from validation.