COMPARISON OF BETWEEN ARTIFICIAL NEURAL NETWORKS AND SOME NON-LINEAR GROWTH MODELS IN ESTIMATION OF NODULE GROWTH OF CHICKPEA PLANTS

This study was conducted under Kirsehir ecological conditions in 2013 and 2014. Cagatay chickpea cultivar was used as material. Plants were sown three sowing times as 42 plants m(2). Plants were monitored after emergency during 5 different time periods and fresh and dry weight of nodules were carefully identified. Measurements were made after emergency, before flowering, flowering period, after flowering and blooming period. Using the obtained data, non-linear model of Richards, Gompertz and Artificial Neural Networks Model were estimated to grow nodules. The successes of the model were compared using different comparison criteria. To conclude, Artificial Neural Networks was successfully predicted nodule growth of cultivated chickpea plants during first and second sowing time. In third sowing time, it was found that the Richards model was resulted in more successful estimation.