Slashdot Mirror

SAS does the same thing, although it is a VM package that I found unusable the only time I tried it before writing my own functions in R instead.

It's not a joke article or astroturfing. He's just using humorous examples of improbable technical solutions to the problem, when of course the real answer is to get more adults involved in helping the kids to learn chess (which is his real point). He's written elsewhere about a K12 chess tournament sponsored by his company:

http://blogs.sas.com/content/s...

Come on.
The article is a joke.
" A chess facilitator brain implant would be wired between perception and cognition. You would just look at the board and know if it is checkmate." Did the original poster not realize this?

It's more along the lines of astroturfing than a joke. The linked article is a blog post on a data analysis company's website. The author is basically dreaming up ways for his company to profit off a minor/nonexistent problem.

Come on. The article is a joke. " A chess facilitator brain implant would be wired between perception and cognition. You would just look at the board and know if it is checkmate." Did the original poster not realize this?

Depends on what you call "mature". I am 44 and work on COBOL a few times a year doing contract work for a couple of healthcare gigs. Mostly maintaining (or porting the programs to SAS http://www.sas.com/) it is actually quite lucrative if you don't mind doing it.

That is your experience and here is mine, I work for a very large hospital in the US (4500+) providers are employed all over the state it is in. The hospital has 2 "Big Data" warehouses (Netezza High Capacity Appliances http://www-142.ibm.com/software/products/us/en/ibmnetehighcapaappl/). My current job is working in the IS Analytics department where we are in the process of consolidating/porting thousands of legacy COBOL programs to SAS http://www.sas.com/. Along with this we also do ad-hoc reporting for providers/users. The amount of data collected at the hospital from other systems and then funneled into these machines is staggering and runs 18/5 right now. If you ever had to deal with QM (Quality Measures) you know the kinds of data that is collected and stored.

I can arrange a tour of the 300+ acre SAS Institute headquarters here in Cary, NC if you would like. It's beautiful - and SAS was Fortune's #1 best place to work in the USA two years in a row - last year and this. I think it's a cool place to work! http://www.sas.com/company/

Because the public sector has very little time to handle FOIA requests and they sometimes cost more money to complete than I'm willing to pay (usually because they don't do much of their own data work in-house and have to call on a contractor to do it for me), I use their websites to glean the data I want.

Last week I gave a talk about using SAS to do screen scraping and then perform analysis on the data of jail inmate registries and level 3 sex offenders in MN. I have dashboards of the data available on my website and as I mentioned in my presentation it has even been used to help one county avoid what could have been a serious privacy issue.

So while there are any number of pitfalls to screen scraping (not understanding the meaning of the data and trends, being fed incomplete or purposefully incorrect data, or even being banned outright) screen scraping can be great for learning about and reporting on the public sector when they are physically or financially incapable or simply unwilling to do it themselves.

SASS would be a Trademark violation.

It should have said:
"If you're in the market for a new job in the US, Fortune has just published its list of 100 Best Companies to Work For in 2010 in the US.

On their own website, SAS admit this is only valid in the US.
http://www.sas.com/jobs/corporate/index.html

I wonder how many holidays US-employees get, how flexible their hours are, how good their coffee is and how much they get paid.

What a nerd.

Have you seen his blog?

It's full of charts and graphs. Not ones like this. Real ones.

... apps like SAS, SPSS, ...

For what it's worth, you have SAS for Linux already. They don't make the product easy to find, and there are a lot of deadlinks, but it works well -- if you can get it. Also, SPSS for Linux has been available for a while, too.

So you can let the researchers work on whichever platform makes the most productive, be it OS X or Linux. But the real boost is for the system maintenance team. Getting rid of the last of the Windows cruft from the LAN means that all those hours spent coddling and repairing M$ junk can be instead directed towards improving the computing environment. Providing better service is a better goal than just trying to keep things running long enough to get work done.

I think you mean Oct 31st, 2010. But yes, in my environment there are several third party vendors that still certify RHEL 3 as the only Linux platform their product will run on (yes, I'm looking at you SAS). RedHat has done a pretty good job of maintaining this release. Just this past weekend I installed it on a new HP Proliant G5 server. I was pretty skeptical that the drivers for the array controllers and NICs were going to work without a lot of dicking around, but RHEL3 Update 9 had everything it needed built in.

iPods, MySpace, Facebook, texting, gChat, Wiis, Game Boys, PlayStations, and, God, especially YouTube all have something in common: using them doesn't make you smarter. Let's face it, the solution is to put down the toys and pick up the books. There's just no way around it. Other countries (China, India) seem to understand this, which is why the admit list for most graduate programs now reads like a Badminton World Cup roster. I look at the study habits of the average American middle schooler and shudder. The future is not bright.

Of course, I would expect no more informed a comment from the guy responsible for the single worst piece of software ever created.

It looks like many quantitative applications are currently not going to work on Vista, at least for now. Major statistical analysis, data mining and Geographic Information Systems tools that don't run on Vista include:

SPSS, SAS, MATLAB and SAP and ESRI ArcGIS

Eh, this is no big deal, right? I mean, who really wants to know about facts and numbers? Especially when you are using a *computer*.

That's a perfect example of what I'm talking about. I worked on something very similar once, as a consultant for a big investment bank, where we had to analyze tick data. I can't believe anyone would do such a thing in Excel. Even if you have a big in-house library built up, Excel's data management capabilities are lacking. You should check out Stata or, god forbid, SAS (if you have a humongous dataset.) I've never understand the prevalence of the Office suite among MBA & finance types when there are so many better programs out there and y'all clearly have the money to spend on them :)

The SAS compiler is a Windows / unix hosted cross-compiler for the mainframe. It's not cheap, and we still won't buy it at work, but it's supposed to do a significantly better job than the IBM compiler / LE.

Ah, Going public. The excitement of Stock Options and being traded on the stock exchange. Everyone thinks paydays will get bigger and the company with thrive and grow.

In reality, what happens is that you are know answerable to the will of mysterious stock holders. You start learning a new phrases and vocabularly like, "shareholder equity", "IPO", "Sarbanes-Oxley", "vesting period", "we must make decisions that increase shareholder value", and "the purpose of stock isn't to make employees rich."

Soon after the IPO, raises and bonuses shrink. Healthcare gets slashed and perks vanish away. Why? Because executives who are now accountable to shareholders rank their company vs. competitors and create a scorecard. Suppose the shareholders were to find out that your CEO was paying better bonuses to employees than the industry standard. He might have to answer for that on an earnings conference call or meeting with the mysterious shareholders. Executives however always want raises, bonuses, perks, and cheap stock no matter what kind of job they do. Just ask the idiot running GM into the ground. He should be well compensated no matter how poorly the company performs.

I think Google thought they could go public and still maintain control of the company, but it looks like they are careening out of control. The absolute best thing that could happen is for Google's stock to crash, then have Google buy all the outstanding shares and convert back to a private company.

There are still some really great privately held technology companies like SAS where life is good for employees. Am I bitter? Sure, I went through the whole IPO process and watched as executives were rolling in cash while they sold stock for which they had paid a mere $.01 per/share. Meanwhile, I had to hang onto my stock and stock options for a vesting period while the price plummeted and they all left to go find another company to rape and pillage. Does anyone know of a situation where going public was actually good for a company and it's employees?

Sure you can.

In addition to that, what software company (Like Great Plains, People Soft, SAS)

Well...MS now owns Great Plains.
SAS? Install Center: SAS for Linux®.
PeopleSoft? Owned by Oracle, who does support Linux.

Have you tried SAS?

Glad to see the above post as being mod'd up. Its not only the cell phone industry, its also the ILEC (Incumbant Local Exchange Carriers).

Capacity has to be considered on few sides:
1) Switch (telco switch, matrix)
2) Transport to and from the switch.
3) The actual signaling links that control the set up and tear down of calls.

The following link illustrates utilization / traffic analysis; its also a brief primer on a telco network.

http://support.sas.com/documentation/onlinedoc/its v/cmgpbx.html

Keep in mind that the tables only show how many trunks are needed to support a flow of traffic. The actual switches have their own limitations (IE: 500,000 calls per hour). For a quick reference: a decked out 5ess goes for about 10 Million (fully equip'd).

Disclaimer: Fully equip'd references that there is no more capacity within that switch to grow beyond that point. Now the actual cards that populate the switch vary in cost. The 10 million mark is reference to a local EO switch (end office).

Headline: Brains More Like Neural Nets Than Traditional Programs

Who woulda thunk it.

ftp://ftp.sas.com/pub/neural/FAQ.html%23A2

'Most NNs have some sort of "training" rule whereby the weights of connections are adjusted on the basis of data.'

Insert joke about the 1980's (or 60's/50's/40's) calling). Somehow I don't think Norbert Weiner would be the slightest bit surprised.

-Tupshin

Isn't it time we all realize that outsourcing is mostly a buzzword? For sure it's a buzzword that cost people their jobs but what is new about that? "Synergies" anyone? (The buzzword which cost me my job). It's the same old PHB bullshit and costs the shareholders losses in the long term.

The PHB eagerness for outsourcing has already begun to bite their ass and the tide is slowly turning into a more balanced view of the real cost of outsourcing for the company. Companies with PHB's are grossly inefficient anyway and doomed in the end no matter what (just because it has "worked" for 50 years doesn't meant it will do so for 50 more).

Just another small bump in the development of the truly global economy. More and more people realize that the PHB's are the problem, as an alternative to PHB's look at the way SAS.com works http://www.sas.com/corporate/worklife/.

Just in case anyone is interested in real, proven data mining software... Check out both SAS and PolyVista. PolyVista is capable of performing some remarkable visualizations that allow you to explore large data sets in ways that just are not possible with traditional OLAP tools.

Then (during the "bubble", yes perks were particularly in vogue). Some of the 90s perks were ridiculous. Netscape was famous for many things, and infamous for some of their perks - onsite free sushi bar, roving free masseuse, etc.

DotComs were offering Ferraris to those who could recruit the most talent. Everyone who was anyone offered stock options.

When the bubble burst, much of the madness was finally seen as madness, and it all went away. That gave many existing companies leverage to take away benefits - "You're lucky to have a job!". Yes and no.

I had a friend who was an attorney for Tandy Corporation (Radio Shack). Tandy paid their attorneys ridiculously low salaries (as in $30k/yr for a real estate attorney). When I asked him what the hell was wrong with them, and why they thought that was appropriate, he told me their response: "These guys are just going to come here for a couple of years and leave anyway, so why should we pay them reasonably?" Duh! Naturally, anyone with talent will move along. That's true in IT as well, and options do still exist. Maybe they involve moving to a new city, but they exist.

Some companies have been doing right all along, and they are rewarded with fierce loyalty and very good productivity. SAS Institute, in Cary, North Carolina, has been providing stellar perks for years. They've remained private, and thus avoided the Quarterly Earnings per Share death-cycle. Imagine if your company had benefits like theirs.

Other companies could be like SAS if they weren't public, and if their leaders understood what some perks could do for their productivity and employee loyalty.

Then (during the "bubble", yes perks were particularly in vogue). Some of the 90s perks were ridiculous. Netscape was famous for many things, and infamous for some of their perks - onsite free sushi bar, roving free masseuse, etc.

DotComs were offering Ferraris to those who could recruit the most talent. Everyone who was anyone offered stock options.

When the bubble burst, much of the madness was finally seen as madness, and it all went away. That gave many existing companies leverage to take away benefits - "You're lucky to have a job!". Yes and no.

I had a friend who was an attorney for Tandy Corporation (Radio Shack). Tandy paid their attorneys ridiculously low salaries (as in $30k/yr for a real estate attorney). When I asked him what the hell was wrong with them, and why they thought that was appropriate, he told me their response: "These guys are just going to come here for a couple of years and leave anyway, so why should we pay them reasonably?" Duh! Naturally, anyone with talent will move along. That's true in IT as well, and options do still exist. Maybe they involve moving to a new city, but they exist.

Some companies have been doing right all along, and they are rewarded with fierce loyalty and very good productivity. SAS Institute, in Cary, North Carolina, has been providing stellar perks for years. They've remained private, and thus avoided the Quarterly Earnings per Share death-cycle. Imagine if your company had benefits like theirs.

Other companies could be like SAS if they weren't public, and if their leaders understood what some perks could do for their productivity and employee loyalty.

If you do know of any, please, PLEASE let us know!

Sas. Now you know.

60 minutes did a piece on this company (you can read Sas's blurb on it here). They are hiring programmers. Nothing below a master's degree and Ph.D. preferred. But if you have one of those and want to work 35 hours/week with competitive pay and obscenely generous perks, you should apply.

If you do know of any, please, PLEASE let us know!

Sas. Now you know.

60 minutes did a piece on this company (you can read Sas's blurb on it here). They are hiring programmers. Nothing below a master's degree and Ph.D. preferred. But if you have one of those and want to work 35 hours/week with competitive pay and obscenely generous perks, you should apply.

This group produced three notable results:

• A specification, which was ultimately submitted to X/Open,
• A declaration that 2**64 bytes is a "bubbabyte", and
• A really cool T-shirt.

I still have my T-shirt -- how about you?

Do you not think that the term "Artificial Intelligence" is more than often misused, especially by so-called "AI" researchers themselves? AI rules generally fall into two statistical categories, classification or regression. Why are they not known as statistical functions then? It seems to me that computer scientists and engineers are particularly keen to play god and hence the misnaming of mathematical functions. See for example this excellant paper by Warren Sarle, a statistician who maintains the comp.ai.neural-nets FAQ.

• not supporting alternative browsers was losing roughly 10 per cent of possible visitors, which was particularly dangerous for a new banking service;
• not supporting alternative browsers was unethical.

It strikes me that the 'ethical lever' would be particularly powerful when charities are concerned.

The organization in question uses a LOAD of PEECEEs as you put it (and some Macs thrown in for good measure) and they are familiar with CD-Rs (at least in some departments) but they are no more capable of taking data out of DB/2 on an ES9000 and getting it into a flat ASCII CSV file on CD-R than we are...seeing as we have no ES9000.

I would think that since they have the mainframe(s) and they have tools like Essbase and SAS and they have the middle-ware gateways to connect them, they could extract the data they need from one of their query tools and burn it onto CD-R. But they don't know how...and they already know how to dump it to 9-track.

That said, if you have the time, implementing it yourself is a great way of learning the algorithms. My PhD topic is neural networks, and I 've always found that implementing a particular architecture myself gives me the highest level of understanding and satisfaction.

That said, if you have the time, implementing it yourself is a great way of learning the algorithms. My PhD topic is neural networks, and I 've always found that implementing a particular architecture myself gives me the highest level of understanding and satisfaction.

As of SAS 8 (possibly before that even) the Output Delivery System (ODS) has supported HTML output. Since one of the main things that SAS does (it's a statistical analysis language) is in fact create reports it does this quite well I think. You'll have some disagree but as far as I've been able to tell the only language people like is the one they program in. So you might want to check out SAS. I've had quite a bit of luck with it. The URL is http://www.sas.com or you want to click http://www.sas.com. You can also contact me via my webpage if you want some more info on it since I (and my employer) consider myself to be good at it.

This this the most ridiculous post I've ever read. As many replies have already explained, a Multi Layer Perceptron (one kind of NN) is an universal approximator, that is it can be used to model a mapping from one set to another, thanks to the data (a set of input-ouput associations).

There are other ways to obtain universal approximation (which means that basicaly any regular mapping can be approximately represented). NO method can be used to give correct EXTRApolations. This is not possible. Period.

Now, regarding INTERpolation of the data, Barron as demonstrated in 93 (see Universal approximation bounds for superposition of a sigmoidal function in IEEE transactions on Information Theory volume 39 number 3, pages 930-945) that MLP are more efficient than any other methods. This means that they use LESS parameters than other interpolation methodes (such as spline, kernel regression, etc.).

So, the post I'm answering to is bullshit. I strongly advise posters to read the NN FAQ before posting ridiculous claims.

Summit?! Ah ha ha ha!
Try Large File support. Oh contraire. You're wrong I'm afraid. The patches are named after the Large File Summit. See http://ftp.sas.com/standards/large.file/ and there's no need to be rude :)

The example of 'prior art' that comes to mind most readily is in SAS.
There is a variable you can set that allows the system to institue windowing for two digit dates based on whichever date you choose. This 'feature' is at least 3-4 years old (I believe) and may in fact be even older.

Perhaps there should be some deliniation of specialty within the USPTO so that the people aproving the patents are required to have some background with the given field for which they
are aproving?

Firstly the link in one of the other posts didn't seem to work so try this.
Secondly which SAS programming language don't you like ? Base SAS ,SCL, webAF or webEIS.
I agree that SCL syntax can be a bit annoying at times but the next version of SAS V8 to be released early next year has much better syntax more like c++ or java. WebAF is just java with a lot of extra classes added to it and a ide so you know what to expect there.

But the main reason for using a product like SAS is that you don't have to rewrite all the statistical and analytic back end procedures. However if you don't like the front end there is a standard server for Open OLAP server available from sas as well as several different web front ends.

For more info check
SAS OLAP
Cognos OLAP
Oracle OLAP

An aside SAS is releasing htmsql 2.0 for Linux as well as all its standard platfroms on tuesday. Does this mean that all of SAS is to be ported to linux ?

Grem

Firstly the link in one of the other posts didn't seem to work so try this.
Secondly which SAS programming language don't you like ? Base SAS ,SCL, webAF or webEIS.
I agree that SCL syntax can be a bit annoying at times but the next version of SAS V8 to be released early next year has much better syntax more like c++ or java. WebAF is just java with a lot of extra classes added to it and a ide so you know what to expect there.

But the main reason for using a product like SAS is that you don't have to rewrite all the statistical and analytic back end procedures. However if you don't like the front end there is a standard server for Open OLAP server available from sas as well as several different web front ends.

For more info check
SAS OLAP
Cognos OLAP
Oracle OLAP

An aside SAS is releasing htmsql 2.0 for Linux as well as all its standard platfroms on tuesday. Does this mean that all of SAS is to be ported to linux ?

Grem

You have to train a neural net. Don't ever underestimate the time, difficulty, and fragility of this step. In order to use a neural net, you have to use a very large data set to intialize it (the dataset tends to grow exponentially as the complexity of what you are trying to do increases). You have to pick the right dataset, which can be extremely difficult if others haven't already figured it out.

You have to train a neural net. Don't ever underestimate the time, difficulty, and fragility of this step. In order to use a neural net, you have to use a very large data set to intialize it (the dataset tends to grow exponentially as the complexity of what you are trying to do increases). You have to pick the right dataset, which can be extremely difficult if others haven't already figured it out.

There are already many companies that just don't
sell their software, they only rent it. For
example, the SAS
institute sells licenses which always terminate...I don't know that you can buy a license that doesn't terminate.

Among commercial software, this isn't always SUCH a bad idea, (if you're of the opinion that commercial software is a good idea in the first place) because some products are "cutting edge" and you really wouldn't want to buy the thing because in two years you'd have to buy it again to get some vital functionality.

It looks like the developers of this distribution are not aware of SAS.

There even is an effort to get SAS ported to Linux.

-- Jochen