Well, Machine Learning doesn't exclude the use of Semantic Tools like Ontologies. You can still use them to gazeteer your ML indexing process, inference over the Ontology hierarchy etc... Both aren't really mutually exclusive. However, I do think the idea of everyone annotating their webpages semantically is never going to take off. The closest thing we have successfully achieved on the interwebz in that sense is WikiPedia.
No you didn't:) It was a valid argument.
However this semantic enhancement requires a couple of things: the model (ontology) must be defined by consensus. A model is by definition an incorrect representation of reality. Hence even with a manually crafted model ontology, it still won't be 'exact'. If you apply this on big medical ontologies, you're really in trouble, as they may have hundreds of thousands of concepts. So this is the ontology part. Next you have the actual semantic annotation part of the document where you put actual trust in the annotator that his knowledge of the ontology is perfect and he's doing a good job of annotating the document. This requires plenty of training.
To clarify, I don't think statistics (ML) would give a better model than an 'exact' - manual - model. I was more speaking in the sense of a 'good enough' system which is also scaleable.
I think there are two reasons why the whole rdf(s)/owl annotated web pages never really gained traction. First of all it's hard work if you have to do it manually, but most content management systems now offer some kind of key word adding feature though. The second reason, IMO, is that the current Big Data and Machine Learning techniques (and more computing power / persistence media / bandwidth than 15 years ago when the whole rdf/owl thing took off) trump the whole categorization and knowledge extraction / data mining process anyway.
As Socrates once said around 500 bc : "Our youth now love luxury. They have bad manners, contempt for authority; they show disrespect for their elders and love chatter in place of exercise; they no longer rise when elders enter the room; they contradict their parents, chatter before company; gobble up their food and tyrannize their teachers."
How about people that took the test with a linux machine? That's a stat available from the browser info as well, no? Didn't check on that now, did they?
The same happened to Marie Curie (you know, the only person to date to win two nobel prizes in different categories and the first to win two alltogether). She was not able to enroll in higher education in Poland because it was not accepting women. Luckily she went to Paris...
For the 'harder' coding silence is definitely golden. But for the monotone 'this is the boring part' coding, just give me a couple of Tool CDs, they are great for keeping focus.
This time it's different though.
Because of the Arab Spring, Western Muslims started to be more engaged around freedom in the Arab countries.
The civil war in Syria (where at first freedom fighters started fighting against the Assad government) succeeded in actually engaging young and influenceable Western Muslims on the field who struggled with their mixed Western and Muslim identity.
ISIS was very successful in utilizing that momentum to draw more young Western Muslims into their ranks later on and many tens of thousands European young muslims (many of them under-age) are now fighting with ISIS. This is very worrisome because these can get back into Europe since they have passports and under the Visa-Waiver program could just board a plane to the US for example.
In that sense, this is now a global conflict. It's not just some tribes fighting against each other and the US intervening to keep control of the oil. It's an exodus of young people who are getting brainwashed and are ticking time-bombs when/if they get back to their actual Western home country.
I don't know exactly which CS curriculum these people went through that didn't have to "work hard enough to understand something", but I'm pretty sure if you have to take (distinct) math classes like Differential and Integral Calculus, Abstract Algebra and Group Theory, Linear Algebra and Geometry; and next to that some other fundamental CS theoretical classes including Lambda Calculus, etc... all in the first year of your CS curriculum (which at least was the case here); you probably encountered a couple of hard nuts to crack.
Add to the mix in the later years a couple of more "applied" CS courses like compilers, interpreters and parsers; and of course; what everyone here is referring to as "good programming" some "easier" classes about Software Engineering, OO programming, Design Patterns and Databases, Networks, and that's basically what a CS curriculum looks like here (except from some possible AI, Machine Learning, Management,... courses).
Of course, this still doesn't guarantee you a good Software Developer/Engineer (passion and motivation is very important), but the basis is as good as it gets.
So, Again, unless you guys have some university level CS curriculum where you graduate building a website in PHP, or unless these Javascript weenies actually had no CS background at all, I'm a bit weary of your statement.
Mind you, I hold EEs in high regard and would definitely hire even if not up to par with latest Node.js FizzBuzz framework.
> The biggest problem in the social sciences isn't their practices, it is that their findings are inherently political
No, the biggest problem in the social sciences is their (often) feeble understanding of Statistics and the heavy publication bias towards positive papers (where significance was found, as opposed to negative papers supporting the null-hypothesis) at conferences and in journals.
If you really don't have a driver's license (as opposed to not having it on you while being pulled over) in Belgium, you will definitely go to police court and probably get a somewhat big fine ( at least 1200 euro). However, you won't get any jail time unless you ran over some school kids.
Belgium is rather lenient considering jail time though, other Western European countries might have you spend the night in jail as well, but probably not get you convicted to actual jail time.
If, in Belgium, you are convicted to jail time but that jail time is less than 3 years, you usually don't have to do them and may get an electronic ankle bracelet instead.
Crap in -> Crap out.
But seriously, don't the common job sites do this already? I'm pretty sure a Monster; GlassDoor or LinkedIn already have departments working on better matching algorithms ( using NLP, supervised or un-supervised machine learning).
I have the same setup however didn't tinker with it for dirty power offs. Never had an issue with corrupted files. Of course usually when I turn the TV off, the pi is at bash, so not really doing anything.
I always liked the HIV test example.
If you have a test that is 99% accurate, you would think that's a pretty good test.
However, if only 1 in 1000 people actually are HIV positive, this means you get 10 false positives per correct positive.
So that's not a very good number falsely claiming people have HIV 9 times out of 10!
Actually, even 99,9% would be bad, since that means you're wrong 1 time out of 2.
When an event emits at a particular wave-length; that event probably emitted at other wavelengths as well. Since different wavelengths travel at different speeds, it's still possible to observe other data from the same event a bit later.
Has anyone looked at the graphs and the "linear correlation" between RME and the "collective intelligence" from the study?
There's all kinds of wrong in there. First of all, looks like the dots of the study show a - very scattered - vertical pattern, with actually the best teams seeming to have a rather average RME (higher end though).
Also, who says there isn't a correlation with intelligence in general and RME? Seems to me people "who care" or "pay better attention" will be better at RME as well.
And what's the task to be solved? Apparently seems to be a sudoku puzzle. If you don't really know how that goes to begin with, you're already at loss (even if you're smarter).
Slashdot Mirror
Well, Machine Learning doesn't exclude the use of Semantic Tools like Ontologies. You can still use them to gazeteer your ML indexing process, inference over the Ontology hierarchy etc... Both aren't really mutually exclusive. However, I do think the idea of everyone annotating their webpages semantically is never going to take off. The closest thing we have successfully achieved on the interwebz in that sense is WikiPedia.
No you didn't :) It was a valid argument.
However this semantic enhancement requires a couple of things: the model (ontology) must be defined by consensus. A model is by definition an incorrect representation of reality. Hence even with a manually crafted model ontology, it still won't be 'exact'. If you apply this on big medical ontologies, you're really in trouble, as they may have hundreds of thousands of concepts. So this is the ontology part. Next you have the actual semantic annotation part of the document where you put actual trust in the annotator that his knowledge of the ontology is perfect and he's doing a good job of annotating the document. This requires plenty of training.
To clarify, I don't think statistics (ML) would give a better model than an 'exact' - manual - model. I was more speaking in the sense of a 'good enough' system which is also scaleable.
I think there are two reasons why the whole rdf(s)/owl annotated web pages never really gained traction. First of all it's hard work if you have to do it manually, but most content management systems now offer some kind of key word adding feature though. The second reason, IMO, is that the current Big Data and Machine Learning techniques (and more computing power / persistence media / bandwidth than 15 years ago when the whole rdf/owl thing took off) trump the whole categorization and knowledge extraction / data mining process anyway.
As Socrates once said around 500 bc : "Our youth now love luxury. They have bad manners, contempt for authority; they show disrespect for their elders and love chatter in place of exercise; they no longer rise when elders enter the room; they contradict their parents, chatter before company; gobble up their food and tyrannize their teachers."
How about people that took the test with a linux machine? That's a stat available from the browser info as well, no? Didn't check on that now, did they?
They are all infringing (it's NOT stealing!).
The same happened to Marie Curie (you know, the only person to date to win two nobel prizes in different categories and the first to win two alltogether). She was not able to enroll in higher education in Poland because it was not accepting women. Luckily she went to Paris ...
For the 'harder' coding silence is definitely golden. But for the monotone 'this is the boring part' coding, just give me a couple of Tool CDs, they are great for keeping focus.
This time it's different though.
Because of the Arab Spring, Western Muslims started to be more engaged around freedom in the Arab countries.
The civil war in Syria (where at first freedom fighters started fighting against the Assad government) succeeded in actually engaging young and influenceable Western Muslims on the field who struggled with their mixed Western and Muslim identity.
ISIS was very successful in utilizing that momentum to draw more young Western Muslims into their ranks later on and many tens of thousands European young muslims (many of them under-age) are now fighting with ISIS.
This is very worrisome because these can get back into Europe since they have passports and under the Visa-Waiver program could just board a plane to the US for example.
In that sense, this is now a global conflict. It's not just some tribes fighting against each other and the US intervening to keep control of the oil. It's an exodus of young people who are getting brainwashed and are ticking time-bombs when/if they get back to their actual Western home country.
I don't know exactly which CS curriculum these people went through that didn't have to "work hard enough to understand something", but I'm pretty sure if you have to take (distinct) math classes like Differential and Integral Calculus, Abstract Algebra and Group Theory, Linear Algebra and Geometry; and next to that some other fundamental CS theoretical classes including Lambda Calculus, etc... all in the first year of your CS curriculum (which at least was the case here); you probably encountered a couple of hard nuts to crack. ... courses).
Add to the mix in the later years a couple of more "applied" CS courses like compilers, interpreters and parsers; and of course; what everyone here is referring to as "good programming" some "easier" classes about Software Engineering, OO programming, Design Patterns and Databases, Networks, and that's basically what a CS curriculum looks like here (except from some possible AI, Machine Learning, Management,
Of course, this still doesn't guarantee you a good Software Developer/Engineer (passion and motivation is very important), but the basis is as good as it gets.
So, Again, unless you guys have some university level CS curriculum where you graduate building a website in PHP, or unless these Javascript weenies actually had no CS background at all, I'm a bit weary of your statement.
Mind you, I hold EEs in high regard and would definitely hire even if not up to par with latest Node.js FizzBuzz framework.
Poor Michael Jackson :(
> The biggest problem in the social sciences isn't their practices, it is that their findings are inherently political
No, the biggest problem in the social sciences is their (often) feeble understanding of Statistics and the heavy publication bias towards positive papers (where significance was found, as opposed to negative papers supporting the null-hypothesis) at conferences and in journals.
I sometimes also have found the average Philosophy major to be indistinguishable from an Eliza program.
If you really don't have a driver's license (as opposed to not having it on you while being pulled over) in Belgium, you will definitely go to police court and probably get a somewhat big fine ( at least 1200 euro). However, you won't get any jail time unless you ran over some school kids.
Belgium is rather lenient considering jail time though, other Western European countries might have you spend the night in jail as well, but probably not get you convicted to actual jail time.
If, in Belgium, you are convicted to jail time but that jail time is less than 3 years, you usually don't have to do them and may get an electronic ankle bracelet instead.
Crap in -> Crap out.
But seriously, don't the common job sites do this already? I'm pretty sure a Monster; GlassDoor or LinkedIn already have departments working on better matching algorithms ( using NLP, supervised or un-supervised machine learning).
It probably is!
People love to believe stuff like that. That a shot of vitamine Z would cure some serious disease.
I have the same setup however didn't tinker with it for dirty power offs. Never had an issue with corrupted files. Of course usually when I turn the TV off, the pi is at bash, so not really doing anything.
I always liked the HIV test example.
If you have a test that is 99% accurate, you would think that's a pretty good test.
However, if only 1 in 1000 people actually are HIV positive, this means you get 10 false positives per correct positive.
So that's not a very good number falsely claiming people have HIV 9 times out of 10!
Actually, even 99,9% would be bad, since that means you're wrong 1 time out of 2.
1. short Alibaba stock
2. post negative official message publicly
3. ???
4. PROFIT !!!
> A classical computer would have to run for thousands of years to compute the quantum equations of motion for just 100 atoms.
But is it web-scale?
OMG, they stole Bennett's idea for moderating Twitter !!!!
When an event emits at a particular wave-length; that event probably emitted at other wavelengths as well. Since different wavelengths travel at different speeds, it's still possible to observe other data from the same event a bit later.
I'm not clicking that link at work. Best Case it's a RickRoll Youtube link, Worst case it's goatse :p
Has anyone looked at the graphs and the "linear correlation" between RME and the "collective intelligence" from the study?
There's all kinds of wrong in there. First of all, looks like the dots of the study show a - very scattered - vertical pattern, with actually the best teams seeming to have a rather average RME (higher end though).
Also, who says there isn't a correlation with intelligence in general and RME? Seems to me people "who care" or "pay better attention" will be better at RME as well.
And what's the task to be solved? Apparently seems to be a sudoku puzzle. If you don't really know how that goes to begin with, you're already at loss (even if you're smarter).