Yield Prediction in Brinjal (Solanum melongena CV MAHYCO-11) Across Different Growth Stages Using ANN Models

Résumé fourni par la source

An attempt has been made in the present investigation to assess the influence of various biometrical characters across different growth stages in Brinjal crop yield along with various yield attributing characters across four growth stages using ANN models. Results of seedling stage had indicated that plant height, girth and number of primary branches could together predict the crop yield to an extent of 83 % for training set and 62 % for validation of model accuracy. Normalized importance was 0.225, 0.142, and 0.131 respectively. In case of Vegetative stage plant height, girth and number of leaves could together predict the crop yield to an extent of 89 %, for training set and 85 % for validation of model accuracy and Normalized importance was 0.126, 0.098 and 0.133, respectively. Where as in Flowering stage both the plant spreads (east-west and north-south) could together predict the crop yield to an extent of 88 % for training set and 64 % for validation of model accuracy and Normalized Importance was 0.220 and 0.245 respectively. Finally for fruiting stage plant height, number of primary branches and plant spread (east-west) could together predict the crop yield to an extent of 68 % for training set and 64 % model accuracy. Normalized importance was 0.229, 0.134 and 0.227 respectively.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.