e.b.ann.step1: Artificial Neural Network - Model generation
In Vitlik/DA2-17: Image Recognition

Creates the ANN Model from the library "mxnet". To get (back) to the overview of all steps and functions use this link: a.a.main

1	e.b.ann.step1(train_x, train_y, rounds, lr, nodes, batch)

`train_x`	A data matrix with features to train on, rows being objects and columns being features
`train_y`	A vector containing all classifications for the given objects
`rounds`	A numeric value that is the amount of rounds the model is trained
`lr`	The learning rate
`nodes`	A vector containing node amounts per layer, e.g. rep(1000, 2) for two layers with 1000 nodes.
`batch`	The amount of objects which the model is trained on every round