25 comments

[ 162 ms ] story [ 3251 ms ] thread
"Data science is powerful because it maximizes leverage in the presence of finite resources."

Regular old "science" does that too.

Hush. Their ego is trying to feel important.

[Old Adjective] [Old Noun] = New Thing We Can Be Experts At Before Everybody Else!

"DJ'ing is powerful because it maximizes sound combinations in the presence of finite source material."

he's not saying data science isn't regular old science. But processing lots of data points has made modern science much more powerful because of the availability of data.
As opposed to non-data science, where they just make things up out of thin air? I think what you're trying to say is that "data science", as a field, is able to be much more productive these days because of an increased availability of pre-processed data. I'd completely agree! Unfortunately the author's statement does not reflect that idea at all (though I'd be lying if I said I understood what they were trying to get at). At the end of the day all science deals with data, and most data is, primarily speaking, of finite quantity, so I'm not exactly sure what leverage is being described here...
"Employee discharge from eyes, nose, or mouth"

With the US's health insurance system and management habits? Makes sense.

It's a little disappointing that the author didn't go into more detail on arguably the most interesting part: the modeling that they explored. It would be really cool to query Yelp's api to gather more data to factor into the model for each restaurant. Unfortunately, Yelp's api always seemed severely limited to me, it's a trove of rich data that they've kept locked down. Their api only allows you to gather 3 reviews on each business [1] and there's no way to query for data about a user.

[1]: http://www.yelp.com/developers/documentation/v2/business

This is great, thanks. How replicable is your process for other cities that have open data?

Code for America recently worked with Yelp and several cities to standardize restaurant inspection data so that services like Yelp (and perhaps you) can display it nationwide.

http://www.yelp.com/healthscores

I went a really mediocre, boring restaurant yesterday that had a score of 98. Didn't change my mind on it.
The part where the author picks out a subset of "unforgivable" violations is arbitrary and unsupported. It's like "what sounds the grossest" with a bunch of Data Science Analysis layered on top.. Because boy do we nerds love our charts-n-graphs. But the assumptions are faulty.

Improper storage and refrigeration (not on the "unforgivable" list) are undoubtably some of the biggest causes of food borne illness.

I just want to take the opportunity to comment further on the hygiene thing, because it's relevant to public health in a very unexpected way.

The author has such a fixation on hygiene that they arbitrarily elevate it to a special category, above proper refrigeration and storage, beyond all reason and logic. What reason and logic, you may ask? Hygiene's mainly an issue if a food handler is sick or come into contact with a sick person. Whereas improperly stored food will get you ill a very high percentage of the time regardless of the health of the staff. Just do a quick calculation of the odds in your head of which is a more critical issue.

But it's super gross, right, and most people can't help but put it at the top of their mental list of "grossness", even above unrefrigerated meat and eggs (that have been previously refrigerated), which are far more likely to make you sick.

So how does an irrational fixation with hygiene affect public health? There was a study done that compared poor Russian kids with rich Finnish kids living across the border from one another. It found higher rates of auto-immune diseases, allergies and diabetes in the Finnish kids. Similar genetic profiles and all that.

Long story short, the theory is the rich Finnish parents were more obsessed with hygiene and keeping their kids clean, and it actually hurt the development of a normal immune system.

Interesting, right? Moral of the story: mere quantity of data is insufficient to draw correct conclusions. One must also adhere to a rigorous scientific methodology that controls for biases.

Source: http://www.smithsonianmag.com/science-nature/The-Unintended-...

I'm going to echo some of these statements here - I'm an infectious disease epidemiologist, and while the author seems to want to speak with authority

"inappropriate cooling methods, improper food storage" are absolutely things that can and will land you in the hospital.

And things that are give no indication of why they'd be unforgivable. "Improper food labeling or menu misrepresentation" sounds like it could mean anything from serving horse when the menu says beef, to something as benign as saying "with a side of garlic butter" but actually using margarine.
I've noticed this trend before, and I don't even have a hypothesis to explain it. To me, clouds of flies inside the restaurant are off-putting. Yet apparently most people either don't notice, or don't care...?
If the food wasn't great, then somewhere so gross wouldn't stay in business very long.

All of my favourite places to eat or drink are hardly cleanrooms. If I were elderly or had a weakened immune system, that would probably concern me more than it does currently.

I don't fear for my health per se, but the disgust curbs the enjoyment.
Doesn't surprise me, either the problems or which restaurants are the worst. Immigrant entrepreneurs from countries where hygiene standards are far below what we are accustomed to in the west just don't "see" these things, or if they do, don't consider them to be problems.
I recently spent four days in hospital on IV fluids after a bad restaurant meal. Here's what I learned.

1) Most restaurants are primarily in business to make money, not to provide high-quality healthy food.

2) Much of the food served in restaurants is not made there. There may be a chain of 3-4 intermediary suppliers between the raw ingredients and what you are served.

3) Most restaurateurs don't know what is in the food; not because they don't care but because it is impossible to monitor the supply chain.

Having nearly died as a result, I will never again trust my food to someone whose primary motivation is financial gain. Now, if I eat it, I made it.

This "primary motivation is financial gain" describes capitalism in general does it not. So you could extend your distrust to all those business that produce the raw ingredients too. You probably do not grow or raise all the food you consume.

Also there may be financial gain in rarely making your customers ill so that does not automatically make food businesses untrustworthy.

The extent I will frequent a Mexican restaurant is generally inversely proportional to how clean it is (or at least appears).
What about the deliberate mishandling and, or, adulteration of food by disgruntled employees?

Anyone who's seen the move "Waiting" knows what I'm talking about.

What happens in the kitchen ends up on the plate

Seems like, that many times, money involved in a business is directly proportional to filth. Odd.