Shahi, B. N.Nayla Fraz2023-02-242023-02-242022-10https://krishikosh.egranth.ac.in/handle/1/5810194425The present study was undertaken utilizing the records of 567 daughter progeny of 55 sires, distributed over a period of 30 years from 1990 to 2019 in crossbred cattle maintained at Instructional Dairy Farm of G.B Pant University of Agriculture and Technology, Pantnagar (Uttarakhand) to study the inheritance pattern of different economic traits and for the prediction of first lactation 305 days milk yield. The Least squares mean (LSM) along with their standard errors of age at first calving (AFC), first service period (FSP), first dry period (FDP), first lactation length (FLL), first calving interval (FCI), first lactation peak yield (FLPY), days to attend peak yield (DPY), first lactation milk yield (FLMY) and first lactation 305-day milk yield (FL305DMY) were observed as 1268.85 ±16.75 days, 267.51 ± 1.93 days, 91.56 ± 1.86 days, 369.41 ± 6.68 days, 465.80 ± 6.64 days, 13.88 ± 0.17 kg, 39.76 ± 0.78 days, 3294.64 ± 77.93 kg and 2570.74 ± 38.30 kg, respectively.Significant effects of sire were found on AFC, FLPY, DPY, FLMY and FL305DMY. Significant effect of period of calving was observed on AFC, FSP, DPY, FLMY, and FL305DMY while non-significant effect was found on FDP, FLL, and FCI. The effect of season of calving was found non-significant for all the first lactation traits under study. The heritability estimates for the traits under study AFC, FSP, FDP, FLL, FCI, FLPY, DPY, FLMY and FL305DMY were observed as 0.34 ± 0.21, 0.10 ± 0.1, 0.12 ± 0.11, 0.45 ± 0.16, 0.32 ± 0.17, 0.42 ± 0.22, 0.40 ± 0.20, 0.54±0.19 and 0.59±0.20,respectively. The genetic and phenotypic correlations between first lactation traits were found to vary from range low to high. The ranking of sires for estimated breeding values (EBV) was compared for all animals using Pearson's correlations or Spearman's rank order correlations. Spearman's rank correlations between sire estimated breeding valuesranged between 0.05 and 0.78. For prediction of first lactation 305 days milk yield different machine learning models were used viz., multiple linear regression (MLR), random forest (RF), gradient boosting regression (GBR), extreme gradient boosting (xgboost) and light gradient boosting (light GBM) and their comparative evaluation was done on the basis of root mean square errors (RMSE) and coefficient of determination (R2) values suggesting that accuracy and precision of RF, GBR, light GBM and xgboost models were adequate in predicting first lactation 305 days milk yield, but the best results were obtained by RF in both training and testing data, it outperformed other regression models in predicting first lactation 305 days milk yield. Machine learning models can be used in dairy industries for the prediction of milk yield in dairy cattle to increase the efficiency of dairy farms and early culling of animals based on 305 days milk yield.EnglishThesis