Hi,
As per my understanding you are trying to train a network with 5 inputs and 2 outputs. You also have a dataset containing 400 rows of data.
There are various kinds of neural networks, each specialising in the type of data it can learn well. For example, a simple fully connected neural network will be good at predicting output values from given set of input values, like in your case. But a fully connected neural network will have a lot more weights and biases than you requested.
With only 5 weights and 2 bias, the network will hardly be able to learn the dataset. You could refer to the following links to learn more about neural networks.
Alternately you could also Google neural networks and find countless blogs and YouTube videos.
