References
Notes {{word-count}}
Summary:
Key points:
Inductive Bias is the knowledge we built into the model to make it learn effectively.
All such knowledge is bias in the sense that it makes some solutions more likely and some less likely.
Note that we can never fully get rid of this because otherwise there is no model.
Learning and Inductive Bias trade-off
Should we build in more knowledge, or better machinery for learning and scale?