verbleibende Zeit $0$:$33$:$48$

Which statements about LSTM applications are true?

a$.$ Fan et al$.(2015)$ trained an LSTM to predict audio from videos of speeches. Given a muted video recording of a speech, they could accurately recover the sound of the speech.

b$.$ LSTM can be used for protein classification. Hereby, the protein is represented as a sequence of bases $($A$,$ C$,$ G$,$ T$)$ and the labels are either functional or structural properties of the respective protein.

c$.$ Radford et al$.(2017)$ used an LSTM to discover sentiment in Amazon reviews by training it for sequence classification $($good vs$.$ bad$)$ on the review text.

d$.$ "word$2$vec" $($Mikolov et al$.,2013)$ is an example for a word embedding technique.

e$.$ A word embedding space is a real vector space in which words are represented as vectors $($points$)$ which often show meaningful relations to each other. For instance, let $f(*)$ be the $($trained$)$ embedding function, then $f($queen