Write a Neural Network in Explicit Form given number of inputs, number of hidden layers, and levels in each layer

39 Views Asked by At

we recently started learning about Neural networks and I would like to get a better grasp of them (or at least make sure I understand correctly);

I'd like to be able to write out the model in explicit form as described in the Title.

We have equations for the hidden layers and for $Y_i$ and when everything is combined and written out we have:

$$Y_i = \beta_0 + \sum_{j=1}^{m-1} \beta_j \alpha_{j,0} + \sum_{j=1}^{m-1} \beta_j \alpha_{j,1} X_{i,1} + ... + \sum_{j=1}^{m-1} \beta_j \alpha_{j,p-1} X_{i,p-1} + e_i $$

So I understand j is the index of the hidden layer, m is the levels associated with the hidden layer and p is the number of inputs (including the intercept). Is this correct?

So I suppose for a network with 3 inputs, 1 hidden layer, and 2 levels in that hidden layer I should have;

$Y_i = \beta_0 + \beta_1\alpha_{10} + \beta_1\alpha_{11}X_{i1} + \beta_1\alpha_{12}X_{i2}$

I guess I'm still having a good bit of trouble with all of this. Something seems off about what I wrote out. Thank you in advance for any help!