Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
Stan Advertising Webpages
Should do soon: writeup for Sloan. Include this example:
On Jan 6, 2017, at 9:12 AM, Linas Mockus email@example.com wrote:
I saw some posts about using stan for federal projects. If it is still relevant we are using for FDA to determine inter/intra variability of different drugs. The model is a very simple hierarchical with error in variables to incorporate analytical error. In fact I am using Brugs & stan to verify the results. We are considering CU, physical properties, and dissolution (fiber optics every minute). We expect bunch of publications over the next 2.5 years.
We are working on NIH proposal and will use stan for compartmental modeling using covariates such as creatine clearance/BUN/age/etc. Probably will use Torsen for this.
Hi all. We've been collecting Stan examples and Stan testimonials, and now we want to put together some promotional materials.
We have various audiences in mind. In no particular order:
Complete outsiders, nontechnical people ("your uncle" or "your grandma") for whom you'd like to explain, at some level, what is this Stan thing you're working on, what does it do, and what's so cool about it.
Current funders: we'd like to make it clear to them that Stan is really being used, for real problems, that it's making a difference in many areas of science.
Technically-minded people who might be doing statistical analysis of some sort R or Python. Why might they want to consider Stan? What can Stan do for them?
Potential new funders: What's so special about Stan, why would the world benefit from supporting it?
Students or beginning Stan users: What are some of the amazing things Stan can do? What's the best way to develop one's Stan skills.
Serious users/developers: Where are we going with Stan? What problems does Stan work well on; what problems is it not so good with?
Probably we can think of other potential audiences too.
The point is that we currently have lots of materials, but not always a place to point people. Stan's homepage is crisp and clean; I like it. But now I want some "brochures."
A Finnish company Smartly.io is proudly using http://www.smartly.io/blog/tutorial-how-we-productized-bayesian-revenue-estimation-with-stan
A Finnish company Reaktor has the apartment price model http://kannattaakokauppa.fi/#/en/ and the code https://github.com/reaktor/Neliohinnat That service is more like advertising of their skills, and it's not public that they are doing something more elaborate for some customer. Reaktor data science team is using Stan for initial experiments and then the final model can be implemented without Stan.