Question

I'm from the vision world and only worked with pixels from 0-255, ignoring any side effects. My current problem is different, in the way that I cannot rely on the input data.

What my problem is:

I have a number of inputs. Each input is categorical (for now) and optional. For example I have a number of user features, {male, female, [..not given]}, {single, relationship, … , [..not given]}, ..

What I want:

X optional Input-Features mapped to Y Output-Features, with uniform output vales across the features.

I tried kernel pca, different kinds of matric factorization for guessing missing inputs and simple autoencoder networks. By sight, the last two yield "ok" results. Any advice here?

No correct solution

Licensed under: CC-BY-SA with attribution
Not affiliated with datascience.stackexchange
scroll top