Institutional Repository of Vilnius University
Pieninių karvių primilžio prognozavimas naudojant neuroninius tinklus /
Abstract
dc:descriptionThis work contains basic information about the workflow principles of neural networks. It also aims to create neural network models that can predict milk yield in dairy cows for the lactation period from the 3rd to the 305th day in milking based on data collected in the time period between the 3rd and the 20th day. To accomplish this a theoretical milk yield is defined. Subsequently, Wood's parametric model (MAE = 0.985 kg, RMSE = 2.019 kg) as well as a modified version of Wood's model (MAE = 0.588 kg, RMSE = 0.929 kg) created in this paper are applied to the aforementioned yield. GRU and LSTM recurrent neural network architectures are trained to predict parameters of the modified model. Errors are calculated between the true and the predicted yield and the model with the best results is selected (MAE = 3.640 kg, RMSE = 5.816 kg).
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Subačius, Paulius,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210642490&prefLang=en_US