Abstract

Abstract The objective of this study was to explore how linear body measurements are related to body weight and can be used to predict calf body weight using linear and machine learning models. To meet these objectives, a total of 69 Holstein calves from a commercial dairy farm were enrolled in the study from wk 2 – 8 of age. Calves were weighed and linear measurements were collected weekly. Nineteen linear measurements were obtained each week, including: poll to nose, width across the eyes, width across the right ear, neck length (NL), wither height (WH), heart girth (HG), midpiece height (MH), midpiece circumference (MC), midpiece width (MW), midpiece depth (MD), midpiece width across the 13th rib (MW13), hook height, hook width, pin height, top of pin bones width (PW), nose to tail body length, the length between the withers and pins (WPL), forearm to hoof, cannon bone to hoof. These measurements were taken using a commercial soft tape measure and calipers. Using a machine learning approach, models were generated to predict BW from calf linear measurements using Weka software 3.8.5 (University of Waikato, New Zealand) using a 10-fold cross-validation method. Both linear regression (LR) and random forest (RF) models were evaluated. Across all weeks the LR model derived 12 of the 19 traits to fit the BW model (r2 = 0.93). These included: PN, NL, WH, HG, MC, MW, MD, HW, PW, MW13, WPL. The RF model slightly reduced BW predictions (r2= 0.92). The results of this study suggest that linear models built on linear measurements can accurately estimate body weight in dairy calves. These data and models generated are important to further the development of visualized weighing systems for young dairy calves and may be used to accurately predict BW without a scale.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call