The toughest data science interview I ever had

I got bombarded for 45 minutes with theoretical questions:

๐Ÿ”ธ Entropy
๐Ÿ”ธ KL divergence, other divergences
๐Ÿ”ธ Kolmogorov complexity
๐Ÿ”ธ Jacobian and Hessian
๐Ÿ”ธ Linear independence
๐Ÿ”ธ Determinant

Continued ๐Ÿ‘‡

๐Ÿ”ธ Eigenvalues and Eigenvectors
๐Ÿ”ธ SVD
๐Ÿ”ธ The norm of a vector
๐Ÿ”ธ Independent random variables
๐Ÿ”ธ Expectation and variance
๐Ÿ”ธ Central limit theorem

๐Ÿ‘‡
๐Ÿ”ธ Gradient descent and SGD
๐Ÿ”ธ Other optimization methods
๐Ÿ”ธ The dimension of gradient and hessian for a neural net with 1k params
๐Ÿ”ธ What is SVM
๐Ÿ”ธ Linear vs non-linear SVM
๐Ÿ”ธ Quadratic optimization

๐Ÿ‘‡
๐Ÿ”ธ What to do when a neural net overfits
๐Ÿ”ธ What is autoencoder
๐Ÿ”ธ How to train an RNN
๐Ÿ”ธ How decision trees work
๐Ÿ”ธ Random forest and GBM
๐Ÿ”ธ How to use random forest on data with 30k features
๐Ÿ”ธ Favorite ML algorithm - tell about it in details

That was tough!

You May Also Like

Recently, the @CNIL issued a decision regarding the GDPR compliance of an unknown French adtech company named "Vectaury". It may seem like small fry, but the decision has potential wide-ranging impacts for Google, the IAB framework, and today's adtech. It's thread time! ๐Ÿ‘‡

It's all in French, but if you're up for it you can read:
โ€ข Their blog post (lacks the most interesting details):
https://t.co/PHkDcOT1hy
โ€ข Their high-level legal decision: https://t.co/hwpiEvjodt
โ€ข The full notification: https://t.co/QQB7rfynha

I've read it so you needn't!

Vectaury was collecting geolocation data in order to create profiles (eg. people who often go to this or that type of shop) so as to power ad targeting. They operate through embedded SDKs and ad bidding, making them invisible to users.

The @CNIL notes that profiling based off of geolocation presents particular risks since it reveals people's movements and habits. As risky, the processing requires consent โ€” this will be the heart of their assessment.

Interesting point: they justify the decision in part because of how many people COULD be targeted in this way (rather than how many have โ€” though they note that too). Because it's on a phone, and many have phones, it is considered large-scale processing no matter what.