How Models Learn: Training and the Three Learning Types

– Training data, labels, and validation sets using a caries detection example
– Supervised learning with labeled radiographs
– Unsupervised learning for identifying hidden groupings without labels
– Reinforcement learning through trial and error and reward
– How training on 50 radiographs differs from training on 50,000
– Garbage in, garbage out: how mislabeled or narrow training data produces flawed models.