Time Series – Hidden Dependency

Once I have been interviewed at Appnexus – an NYC technology firm – and they gave me a data challenge, which I failed (…cold shower). This was before I knew that technology companies like to challenge potential candidates with puzzles every step of the interview.

I got so upset that I designed my own interview puzzle and sent it to the guy who interviewed me. He never replied, so I don’t know if he solved it or if he thought it was below his dignity (I wouldn’t be surprised – he was supersmart).

You’re given a dataset with a few columns, one of the columns corresponds to time (T) and the other to the metric you’re trying to predict (sum of S). There are also columns A, B and R. The challenge is to figure out a formula which would predict the sum of S for every value in T using values in A, B and R – you can use any construct like SQL aggregate functions on rows (too much information already). It helps to keep in mind that solution has to be exact. The good thing for you – dear reader of this – is that my interviewer never let me know there was a hidden dependency to begin with – whereas you are already equipped with that knowledge from the start. That’s a huge aid.

The person who figures this one out is clearly very intelligent – I would be delighted to speak with you if you send me full solution, explaining every step of the way that got you to it. As usual, there is $$ for reward.

The dataset is in the Excel file and the chart below shows sum of S vs T.

Data_challenge

Advertisements

~ by Monsi.Terdex on November 30, 2012.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
Normal Boy

Nothing out of the ordinary

Data Engineering Blog

Compare different philosophies, approaches and tools for Analytics.

%d bloggers like this: