I am trying to derive the gradient of the negative log likelihood function with respect to the weights, $w$.
The solution is here (at the bottom of page 7).
However, I keep arriving at a solution of
$$\ - \sum_{i=1}^N \frac{x_i e^{w^Tx_i}(2y_i-1)}{e^{w^Tx_i} + 1}$$
Is there a step-by-step guide of how this is done? I can't figure out how they arrived at that solution.