Determination of the Effect of Some Properties on Egg Yield with Regression Analysis Met-hod Bagging Mars and R Application

Demet Canga, Mustafa Boğa

Abstract


In the study, it has been demonstrated its use for a data set obtained from layer hens in a hybrid approach obtained by combining BAGGING and MARS. In the study, the data of 2018 of the egg production enterprise in a private livestock enterprise in the Çukurova Region of Adana province were used. In the research, a data set obtained from Lohman breed chickens, who are at an average age of 60 weeks, was used. Earth (enhanced adaptive regression through hinges) and caret (classification and regression training), mda (Mixture Discriminant Analysis) packages were used in R STUDIO program to provide a stronger solution of regression problems in the created MARS and Bagging MARS algorithm. The estimation performance of the bagging MARS technique was evaluated with the goodness of fit criteria by taking the B value of the bootstrap sample number 3. In the study, the effect of temperature and humidity on egg yield, broken / cracked eggs, number of dead animals and feed consumption was investigated using MARS and bagging MARS analysis. While the effect of evening temperature(t3) on egg yield was found to be significant, it was not included in the estimation equation since morning (t1) and noon(t2) temperatures did not have a significant effect. Since the number of broken / cracked eggs and dead animals is less than 5 weeks, these variables are not included in the estimation equation in MARS and Bagging MARS models. It has been observed that feed consumption has a positive contribution in both models.

Keywords


Bagging MARS; Bootstrap; Aggregating; Lohman breed; Temperature; Egg yield

Full Text:

PDF (Türkçe)


DOI: https://doi.org/10.24925/turjaf.v8i8.1705-1712.3468

 Creative Commons License
This work is licensed under Creative Commons Attribution 4.0 International License

ISSN: 2148-127X

Turkish JAF Sci.Tech.

Turkish Journal of Agriculture - Food Science and Technology (TURJAF) is indexed by the following national and international scientific indexing services: