Did you ever try extending it out to other methods of probability estimation other than the forms of regression? I have only skimmed your excellent article, but I think you are first calculating the average probabilities from a regression model and then minimizing the loss to calculate Harville corrections for place and show markets? Is that correct or am I missing something here? I guess I am curious if there has been any improvement on using regressions for combining the various initial odds as I don't really follow the literature anymore.
Yes! There have been big improvements since then but they are beyond the scope of the post. I just wanted to reproduce the calculations in the paper using PyTorch.
Bill Benter subsequently replaced the multinomial logit model with a multinomial probit model, which assumes Normally distributed errors rather than errors that follow the Laplace distribution.
This is used by a number of betting syndicates. Notable by David Walsh[0] and Zjelko [1]. They knew of and worked with Benter in HK and adapted his system. One of the things that they did was pay large numbers of experts to watch and evaluate each horse and give it a standardised rating which they then used as an extra parameter along with the public odds. Another gambler who also extended that system out was Alan Woods [2], but I don't think he did anything as sophisticated in terms of modeling, concentrating more on the execution side. Regarding the yearly turnover, I have no idea, but billions would be on the low end. From the ATO filings we David Walsh alone was generating 100's of millions of dollars a year in profits, let alone turnover.
Is anyone actively organising against this bill? I feel that ever since the Iraq war protests failed ever time some thing like this happens, people complain a little bit, but don't actually manage to change anything. I was wondering if there are any groups out there that are actively protesting this that I could join, or if not, if any one is interested in forming one? It seems to be an issue that will affect the majority of the readers of HN in a negative way, regardless of your usual political affiliation.
People seem to just express their anger at news facebook pages these days, but are far too apathetic to actually go outside and do something about it.
There also seems to be a growing "anti-complaining" feeling around people's interactions, where it at least appears that a large number of people find it amusing to actively attack those who are highlighting a problem.
Ask someone who was an environmentalist for a couple of decades if shooting the messenager is a new or growing phenomena. It definitely isn’t, and is always at its peak issue makes people feel powerless, and they know that it would take personal sacrifice to make a difference. Few people like being told that rough seas are ahead, and the only way to make it through is with extensive teamwork, compromise, and putting aside petty personal issues.
Of course it’s often enhanced by orgsnizations which benefit from the status quo. For a long time being in favor of EV’s fell into the “tree hugger” category of ridicule and censure, and only when it became possible to adopt the tech without significant personal sacrifice did that change. The idea of organizing people and exercising mass political power is obviously hard, potentially risky, and involves more than making a 5 minute video or paragraph of posting. If people have already chosen not to do that, they tend to resent the people loudly reminding them that there is another way they simply chose to ignore.
Piggybacking of this post, if anyone has any similar resources for best practices in modern C I'd be interested. There don't seem to be any moocs that go beyond introductory material and K&R obviously doesn't cover modern tools and methods.
“C Interfaces and Implementations” by David R. Hanson is what I used when I had a similar concern a few years ago. It’s from 1997, so no, it doesn’t cover the newest C standards, which I understand is your main concern. Though C11 didn’t add too much semantically, for what it’s worth. That said, the book provides great design patterns for writing C in an object-oriented sort of manner, more idiomatic to OOP.
Peter Hintjens was writing “Scalable C” a couple years ago, but recently died, and never got very far unfortunately.
There's the "Modern C" book from http://icube-icps.unistra.fr/index.php/Jens_Gustedt. Very opinionated, but it covers a lot of ground and if you disagree, it gives you a framework for figuring out your own preferences.
My biggest question would be what modern means to you? Does it mean the style of the newest popular C project? The newest standard (C11)? And what devices are you using C for? I'd assume a standard POSIX environment, but you can't be sure with C.
Counter steering tips the bike into a corner. Some form of lateral force from the front tire and gravity from the lean angle are responsible for the bike turning. In modern racing the limit on the speed you can carry through a corner is the force between the tire and the road surface. This is proportional to the size of the contact patch between the tire and the road.
Modern traction control systems have recently surpassed humans in being able to maintain a maximum amount of power to the rear wheel without making it spin out. Previously more lean usually ment a shorter line through a corner but was balanced by the need for accurate throttle control, leading to the rider being tucked in close to the bike with one knee down. With the new systems the aim has become to stand the bike up again as quickly as possible to maximize the contact patch allowing the traction control to put more power through the back wheel and accelerate the bike faster. This has changed the entire style of riding for professional racers, leading them to hang further out from the bike often touching both elbow and knee to the ground, to try and maintain the same center of gravity of the bike and rider but with the bike in a more upright position creating a larger contact patch and allowing the traction control to accelerate harder.
Until the last couple of years, the manufacturers all used their own electronics, but now all the bikes must use Magneti Marelli electronics and must figure out how to use those APIs to control spin and throttle. Last season, Yamaha couldn't get this right with its chassis, and its riders couldn't make the tyres last the race with decent grip.
Yes, MotoGP allows traction control. Like all technical innovations it has led to a change in riders priorities and skills. Some people dislike it, but overall it's not accepted I'd say. Personally I've found it interesting to watch riders adapt there riding style to the new technology.
Yes, MotoGP uses traction control. I can't speak for anyone else, but I don't really care. Like F1, it's a prototype class, so results are a combination of the rider and the machine.
To stop consuming and start creating. Also to actually interact with people online. I haven't really commented online or been social online since the late 90's and I want to try and change that and actually create and interact again.
Wondering if anyone has any experience using lisp for machine learning? I'm aware of mgl[0], but it seems to be abandoned. The lack any wrappers for tensor flow or caffe is also a bit surprising to me. The cliki page [1] is also unhelpful and out of date. Is machine learning on lisp dead or are there projects out there that I'm just not aware of?
Not directly, but have been following various projects over the years..
My take on this is that the people using CL for machine learning have been doing this for some time, and so have their own toolsets; tensor flow is relatively new in that regard, and w/r/t lisp interfacing would entail low-level binary interfacing (and therefore mostly non-portable between implementations) to hook CL code with tensorflow kernels (definately not an expert here on either however).
also, what is popularly referred to as 'machine learning' is in my opinion mostly
only one aspect of the field - e.g. classification neural networks - while lisp can
definitely do this, lisp AI programming (in my amateur opinion) shines more in the realm
of machine reasoning/inference due to the symbolic / dynamic nature of the environment - e.g. constructing a set of reasoning primitives (e.g. functions and facts and decision trees) and a meta-interpreter to reason/infer about external data and walk around a problem space.. also, owing to the dynamic and rapid development nature of the language, likely many people are working with their own prototype/core frameworks, possibly cobbled together from various small bits and pieces of 3rd party code. Also, neural networks have been around for quite a while - what these new frameworks bring to the
table is not so much new core algorithms, but the ability to quickly cobble them together in a more popular/user friendly way, and also take advantage of fast hardware (e.g. GPUs)
as for projects - in the general sense, lisp has been a latecomer to the 'languages with cpan-style trove of public addon modules' crowd, owing in my opinion to the need to support multiple implementations in order for such a project to take hold - so older but yet still quite functional libraries might be around in various hodge-podge repositories which are not standardised but old timers have already included in their own local systems, etc. (see also CMU AI repository)
In the last few years, much has been done in the module space - I would
definately consider 'quicklisp' to be roughly the defacto definitive list of current modules, especially those under active development, since the active community is basically converging on this as a module/distribution platform and so many (most?) active community projects are available as quicklisp modules and included here - and so would probably be the one of the first places to start in checking into for available libraries for any topic.
Also, the best way to 'explore' quicklisp is to install it, and then install various packages and then muck around/explore with the source code that they download into
your environment - the documentation tends to be much less 'external' (e.g. websites) and much more 'internal' (e.g. READMEs, in-tree code examples or unit tests).
Feynman nano (unnamed Infection-resistant materials) definitely ticks both the ambition, problem and market boxes. There is currently a ~$30 billion market for catheters, production at scale of nano-structures is currently unsolved. Combined with the current growth of antibiotic resistant golden staph infections, they would seem to be poised for market dominance if successful.
You don't need to even use the ACC, people for get that the 2005 counter-terrorism act[0][1] has provision for preventative detention without charge, and notably, made it a criminal act to tell anyone that you had been detained. Combined with a rather broad definition on what was terrorism and the ability of police to request information, documents and emails, this act seems to cover all of the functional aspects of the National security letters with even less oversight.
