Three ways an AI learns

Supervised, unsupervised (clustering) and reinforcement learning, test data, overfitting and confidence level.

In het Engels Dit pakket is in het Engels: het speelt wanneer het spel in die taal staat.

Vragen

24, in 4 delen: Supervised · Unsupervised · Reinforcement · Confidence and limits

Hoofdstuk

Supervised, unsupervised and reinforcement learning

Spel

Pixelia

Taal van het pakket

English

Een voorproefje

1. In supervised learning, training examples are...all identicalhidden from the AIlabelled with the correct answerwithout any data
2. In unsupervised learning, data...has no labelsis all falseis always soundis labelled
3. In reinforcement learning, an AI learns through...a dictionarylabelspure luckrewards and penalties