Case Study — Machine learning about sexual orientation?

“All we really know is that a deep neural net can draw a distinction between these two groups [in the dataset] for reasons that we don’t really understand.”

Source: Case Study — Machine learning about sexual orientation?

 

This is a nice summary of some problems with the conclusions made in this research study. I would also suggest that the shape of a person’s face, gay or straight or whatever, is far more affected by the shape of their parent’s face rather than the presence of a gay gene.

Advertisements

2 thoughts on “Case Study — Machine learning about sexual orientation?

  1. “far more affected by the shape of their parent’s face” – This is besides the point, surely? It’s also far more affected by the condition of being human or ape, but that doesn’t itself rule out the possibility of a detectable signal.

    (Also you make an uncharacteristic causal fallacy in implying that a parent’s face has a direct influence on the shape of the child’s face, rather than sharing common cause ;)

    Good article tho

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s