Training Sets

A trained deep neural net model is only as good as the curated dataset that it was trained on. ,It will indeed be biased based on who gathered the data, how it was curated, how it was validated, how it was balanced.

It is likely that following the human example, e.g., of politicians, models are trained that think heavily biased statements are neutral, and vice versa.

In a real world situation, governance, transparency, accountability and traceability of the training data is key. This would include crowd sourced options, so that everybody can contribute to, and maintain it. We cannot necessarily trust the intentions or affiliations of people who (read: lobbied) trained the models.

#Machinelearning #bias #minimizebias #AI #classification #governance


Leave a Reply

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

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

Facebook photo

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

Connecting to %s