8 comments

[ 1.7 ms ] story [ 31.4 ms ] thread
Can someone explain what's really an amazing about this? It sounds as though their using machine learning to predict tomorrow's sales in various categories. Why is this surprising or hard?
Its not really hard to predict normal scenarios. From the article " it helped the company to predict that it would sell more frozen yoghurt in the last two days than it has in the whole year."

For me, that is an extreme outlier and if they are able to predict even that, then there is something special about this algorithm.

It may just be that it was the first hot days of the year...
I hate it when some relatively simple piece of math is described in the most complicated/impressive possible terminology to give the biggest splash.

They used a machine learning algorithm to reduce their food waste. Cool! That's a good thing - but calling it 'atomic scientists algorithm' does not really help anyone understand what the algorithm is doing.

I guess the next step is to describe a formula that involves addition and describe it using rank0 tensors and Einstein notation?

I would think next would be including general reletivity in miles per gallon calculations on cars, or adding the mass gained from moving at highway speeds when finding desired tire pressure. It's always fun to throw those in just to confuse people. If anyone questions it, you can honestly say it is a more accurate method.
so, where can i find more information about this NeuroBayes algorithm?