Stanford on-line Data Mining Courses ($10k for a certificate)

Your approach (result-driven hiring) work well in small organization. In big organization certificates simplifies and standardize hiring process.

I am not saying I support it, I am just saying this is being used and if student aims to work for big company, he probably better invest into this piece of paper.

> I do like the fact that elite schools are offering online certificates, but for $10K there has got to be a great value proposition.

The value proposition is that you have someone else pay for them. This is something that you can have most employers pay for as a continuing education proposition. It's the same way I got my MS without paying a dime.

In Florida, in their state university system, a "certificate" is the replacement for a "minor". This isn't a total replacement, as one can get one's "certificate" issued separately from the diploma (I did this).

SCPD is generally aimed at people with training budget money to spend that need some kind of certificate to show for it. There are a few more (Stanford Advanced Project Management some in financial engineering and risk management, some in bio engineering, IIRC).

The OpenCourseWare classes, which have a lot of spiritual overlap and more delivery system lap with SCPD, are aimed at everyone else.

Oh, the SCPD GPA requirement is not particularly strict, it seems, if you have a professional background in the area or a degree from ``top tier'' school or such. They just don't want to let anyone in who will reflect particularly poorly. (Which you might argue the corporate orientation defeats, but that's a different argument entirely.)

So, being someone in the market to hire someone with Data Mining and AI skills, this: " i feel like there's not much it shows beyond commitment and a focused interes" is very valuable. Obviously there are other ways that may be cheaper, but this is not a bad way.

The other things I look for are evidence of raw smarts and a track record of accomplishment, CS fundamentals, and, of course, personality. (Also, prefer someone with a poor understanding of Football so that they don't upset me in the fantasy league.)

I've used SPSS and SAS for the following tasks:

churn analysis, market basket analysis, sequence analysis, segmentation, anomaly detection, risk management, forecasting

Given a real world problem, algorithms or the mathematics behind it do not provide insight on their own. The only value knowledge of the algorithms provide is a general understanding of the ways in which you can approach a problem. If you can do the mathematics, you can look at how the way you structure your question influences the results you are seeing.

The process of exploratory data analysis is best done in (imho) the style of a Lakatos research programme.

- prepare and clean the data

- explore using many fast methods and charts, come up with some working hypotheses about the data that are important to your client

- select method(s) to test those hypotheses

- perform the analyses

- determine what your results mean in terms of your research goal.

- alter your list of working hypotheses

- repeat [possibly collecting more data]

Obviously the only hard part about this is step five. And unfortunately this is the step that isn't t really taught in my experience. A simple case: Let's say you had a linear regression and you ran it once with 2 variables, got some parameter estimates (a,b) , and ran it again with 3 parameters and got some more parameter estimates (a1,b1,c). If b != b1, what does this mean? If you are using a custom link function (e.g. cloglog or logit), how should you interpret this now? This is where having a deep understanding of the mathematics behind the techniques starts to pay off. And this is the simplest example of a basic regression.

There's almost no competition between the two. Mainly because you can't really put free courses on a resume.

The education is free but the credentials aren't?

Online education isn't any more free to provide than anything else. And not all courses will scale equally well either. For instance, would a data mining course include several thousand cpu hours per student?

One would hope they are working towards a sustainable model for efficiently allocating education services, and doing so will require reasonable fees for some courses.

One thing to consider is that large organizations often time have business units work against each other just as a sort of accident or unclear goals.

The reason for this is that the parts of "Stanford" (an organization comprising tens of thousands of people) involved in each push are basically unrelated.

The paid-certificate program is run by the Stanford Center for Professional Development ("SCPD"), and is a trimmed down version of the online masters program. Costs here are high because courses are usually paid for by an employer investing in employee training; the SCPD program (like masters programs in general) is a profit center.

The db-class, ml-class, ai-class websites (and their video-only equivalents through openclassroom and classx) are being run by the professors who teach the classes, and the goal is education.

There isn't a "Stanford" policy on how online instruction should operate; just different members of a single club working independently.

Likewise, ml-class.com is currently being offered and covers some of the material under discussion.

It's not an entire graduate program. Instead you'll usually see this called a post-graduate certificate program. It's a series of class that area all in the same specific area (in this case 3) that at the end you'll have a signed piece of paper from Stanford that says you took these classes and passed.