thesis1/data/010Normalising Flows.txt

9 lines
354 B
Plaintext

<frame title="Normalising Flows">
<list>
<e>Deep Learning Method, in which the output is normalised</e>
<e>$\int f(x) dx=1 \; \forall f(x)$</e>
<e>Can be used to estimate probability density functions</e>
<e>->Thus useful for AD</e>
<e>$\int f(h(x)) \|\frac{\delta h}{\delta x}\| dx=1 \; \forall h(x)$</e>
</list>
</frame>