Machine Learning

Models, baselines, evaluation, and not fooling yourself.

For projects where the method is the model. Expect hard questions about your baseline, your train/test split, and whether your accuracy number means anything.

Nothing in Machine Learning yet

Start the first discussion here — it makes the channel useful for everyone who arrives after you.

Start a discussion

Channel guidelines

State your dataset size and how you split it. A result without a baseline is not a result.