"id": "b4629e2c",

"metadata": {},

"source": [

"Note that the magnitude of the signal seems to be much larger than $Q$. This is because we have a Guassian process $\\implies$ the signal consists of a sequence of \"impulse-like\" functions that have magnitude that increases with the time step $dt$ as $1/\\sqrt{dt}$ (this gives covariance $\\mathbb{E}(V(t_1) V^T(t_2)) = Q \\delta(t_2 - t_1)$."

"Note that the magnitude of the signal seems to be much larger than $Q$. This is because we have a Gaussian process $\\implies$ the signal consists of a sequence of \"impulse-like\" functions that have magnitude that increases with the time step $dt$ as $1/\\sqrt{dt}$ (this gives covariance $\\mathbb{E}(V(t_1) V^T(t_2)) = Q \\delta(t_2 - t_1)$."

]

},

{