Nothing statistical keeps you from doing this.

I do not see much value to the nonlinearity that a neural network will introduce. However, a neural network might do a good job at capturing interactions, and those interactions could be key. My example here gives an example where you need the interaction to distinguish between outcomes.

Nothing statistical keeps you from doing this.

I do not see much value to the nonlinearity that a neural network will introduce. However, a neural network might do a good job of capturing variable interactions, and those interactions could be key. My example here, for instance, gives an example where you need the interaction to distinguish between outcomes.

Nothing statistical keeps you from doing this.

I do not see much value to the nonlinearity that a neural network will introduce. However, a neural network might do a good job at capturing interactions. Those interactions could be key. My example here gives an example where you need the interaction to distinguish between outcomes.

Nothing statistical keeps you from doing this.

I do not see much value to the nonlinearity that a neural network will introduce. However, a neural network might do a good job at capturing interactions, and those interactions could be key. My example here gives an example where you need the interaction to distinguish between outcomes.