Comparing artificial and convolutional neural networks with traditional models for Genomic prediction in wheat