Let $\mathcal{H}=\{ x\rightarrow \langle \mathbf{w},\Phi(x)\rangle+b : \| \mathbf{w}\|_{\mathbb{H}} \le \Lambda, b\in \mathbb{R}\}$ be a function family, where $\Phi$ is a feature mapping, and $\mathbb{H}$ denotes the feature space, i.e., Hilber space.
How can we bound the Rademacher complexity of $\mathcal{H}$?
I know that when $b=0$, the Rademacher complexity of $\mathcal{H}$ is bounded by $\sqrt{\frac{r^2\Lambda^2}{m}}$, where $K(x,x) \le r^2$ and $m$ is the number of samples. But when the offset $b$ exists, how can we bound this Rademacher complexity?