Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What has happened down here is the winds have changed

andrewgelman.com

Post reply on HN
Print view View original

181–187 of 187 posts

Re: What has happened down here is the winds have changed

#181

Post by bjterry »

cvwright wrote: post #175
bjterry wrote: post #166

Earlier quoted context omitted.

One typical strategy for spam detection is to convert text to a "bag of words" representation[0]. If you take this bag of words representation and hash all the values, then rather than training words like ED, you are training on word number 19213123. The number of these hashed values is smaller than the number of words, just like a hash table, and this generally doesn't harm the accuracy of the machine learning. When…

This is actually a really tricky topic. Things that sound like they should give very good security, often don't in practice. The "hashed bag of words" technique that you describe here sounds an awful lot like some recent attempts at letting legacy systems search on encrypted data [1,2]. We took at look at this recently [3], and it turns out that mapping the word numbers back to the original words is actually a lot mo…

Yeah, I definitely don't think that this would give you mathematically provable security, especially if you are including n-grams which would allow you to chain together sentences combined with a language model. How much that matters in practice depends on the dimensionality reduction. In the context of google I doubt if this is even much of a specific goal, since no matter what they train on, they actually have access to the source text if they want it.

Re: What has happened down here is the winds have changed

#182

Post by eveningcoffee »

zmb_ wrote: post #134
eveningcoffee wrote: post #100

Earlier quoted context omitted.

This was the point. You have to pay the price for the principles and hope that the tide turns one day. If you take your suggested easy route then this day may never come.

Are you a researcher and following those principles? Because it is very easy to demand other people pay the price for what you believe are the right principles. The reality of the way research funding and academia is set up (at least in Europe) is that the price you pay is likely your career. You can go through all the work and effort to follow those principles, and all that will happen is that the politicians will g…

I did not demand anything. I just explained that when everybody complies, nothing will probably change.

Re: What has happened down here is the winds have changed

#183

Post by effie »

vacri wrote: post #148
Animats wrote: post #116

Psychology as a science has far fewer reliable results than it claimed it did. It's not that the winds have changed. It's that the field was broken. Realizing this is a good thing. It's going to set back careers and reduce funding, though. Next, economics? "Science is prediction, not explanation." - Fred Hoyle

It's not so much that the field is broken, as that the field is ridiculously hard. Remember that every time you see someone on HN talk about A/B testing; employee relations; the effect of open-plan offices; UI and UX; how to extract payment from a recalcitrant client; anything to do with politics; why real-estate costs are high and their effect on a neighbourhood; or even right down their IDE's colour schemes, they'r…

> ... yet considerable effort is spent trying to figure out the Big Bang - and it's not like you'll find anyone claiming those efforts are "not science".

Good observation. I thought there was a context to Hoyle's statement, so I checked the Wikipedia:

Hoyle died in 2001 never accepting the Big Bang theory.[25] "How, in the big-bang cosmology, is the microwave background explained? Despite what supporters of big-bang cosmology claim, it is not explained. The supposed explanation is nothing but an entry in the gardener’s catalogue of hypothesis that constitutes the theory. Had observation given 27 Kelvins instead of 2.7 Kelvins for the temperature, then 27 kelvins would have been entered in the catalogue. Or 0.27 Kelvins. Or anything at all." — Hoyle, 1994[26]

https://en.wikipedia.org/wiki/Fred_Hoyle#Rejection_of_the_Bi...

Re: What has happened down here is the winds have changed

#184

Post by edem »

danso wrote: post #149
edem wrote: post #146

After reading the first paragraph I still only have a vague concept what this wall of text is about. Can somebody summarize it with a TL;DR?

Psychology must get used to the idea of transparency and to the reform made possible by how the Internet (as a broad term to generally describe the current change in communication) has made it easier to scrutinize and challenge scientific results.

Thanks!

Re: What has happened down here is the winds have changed

#185

Post by squozzer »

ktRolster wrote: post #66
Pxtl wrote: post #52

Earlier quoted context omitted.

While I am sympathetic to people concerned with the amount of scientific trash talk online, I have to agree with the basic thesis: Evidence-based policy is important. The alternative is gut-feel based populist wingnuttery. To stand against Trumpism, evidence-based policy needs to be above reproach. The correct way to be above reproach is to be flawless, not to silence informed criticism.

This is the counter-argument to evidence-based policy: http://www.bmj.com/content/327/7429/1459.long The point being that not everything we know as true shows up in a study.

LOL. Reminded me of a friend who skydove/skydived. He read a magazine about the sport, which had sprinkled throughout concise descriptions of skydiving accidents.

They had one thing in common - cause of death:impact.

Re: What has happened down here is the winds have changed

#187

Post by bumbledraven »

Animats wrote: post #116

Psychology as a science has far fewer reliable results than it claimed it did. It's not that the winds have changed. It's that the field was broken. Realizing this is a good thing. It's going to set back careers and reduce funding, though. Next, economics? "Science is prediction, not explanation." - Fred Hoyle

Fred Hoyle was wrong. Science is about explanations. David Deutsch explains why in _ The Beginning of Infinity: Explanations that Transform the World _.

The catch is that the explanations have to be "hard to vary"; that is, they have to be non-arbitrary. That's why a field like psychology fails while fields like math, micro-economics, biology, and physics succeed.

Post reply on HN