Virtual StatsPD@Waite meeting

Every month, the professional development meetings of statisticians and data scientists at Waite, known as StatsPD@Waite, bring together specialists in various aspects of data sciences in agriculture from Waite, Roseworthy and Adelaide.

Please join us for the next StatsPD@Waite meeting where Mario Fruzangohar (University of Adelaide) will present on Genomics Selection in Plant Breeding: Linear Models vs Deep Learning Models.

Please note that the StatsPD@Waite meetings are recorded. If you have a question to the speaker but would rather not be recorded, please send me your question via chat during the meeting and I will ask it on your behalf. 

Please email Beata Sznajder for details of the Zoom meeting.

Genomics Selection in Plant Breeding: Linear Models vs Deep Learning Models

Mario Fruzangohar (University of Adelaide)

Linear Mixed Models (LMM) have been extensively used in plant breeding programs to predict the genetic performance of varieties using approaches such as Genomic Selection (GS). There is now an opportunity to improve the accuracy of GS through modern Machine Learning (ML) methods such as Neural Networks (NN). In this talk, we will discuss the statistical background of LMM and NN approaches in the context of GS in plant breeding. To illustrate the approaches, a large diversity wheat panel data set from Australian Grains Technologies (AGT) was used to conduct GS with an LMM and a cutting edge Bayesian Dropout NN. Prediction accuracy of the LMM and ML approaches was then compared and the Bayesian Dropout NN was shown to slightly improve GS accuracy.