W1 Ch1

[JSchuurmans]

Today, most computer programs -> Most computer programs

Explain unsupervised more, you only give an example. Unsupervised Learning = Trying to find hidden structure in unlabeled data. (Clustering/Hidden Markov Models/Self Organized Maps)

Semi-supervised Learning, you have some labels. Class of supervised learning tasks and techniques, where you have some labeled data, but the majority is unlabeled. (Different from reinforcement learning)

Name Generative models in a different list, yes it is a subfield of ML but the previous are ways of learning. Generative (together with Discriminative) are model types.

GAI, 'computers that are as smart as humans', too simple. At least mention 'good at more than one specific task'.

GAN, the added value of GAN is not just that it creates examples (bootstrapping does the same). It generates examples where the NN needs improvement.

Move 'Returns free of market risk.' up one sentence.

'(although many companies...' deserves to be its own sentence.

'... and then shorting a market index.' not necessarily https://www.investopedia.com/terms/h/hedgefund.asp

Because the competition... -> The competition.... Because at the end.... -> At the end.... (Three times because & Although officially you can try not to start a sentence with Because)

'especially those with many layers of neurons,' -> leave this out, people don't know yet what a layer of neurons is.

Deep neural networks and deep learning is something different. You are describing deep neural networks. Delete 'deep learning'.

'complex statistical patterns, such as the statistical pattern' delete first 'statistical'

'difficult to describe statistical rules' they can be complex logical rules as well (Regular Expressions for example)

'But it is not only...' -> It is not only...

[JannesKlaas]

All issues addressed, will be updated in next push