opensandwich

Perhaps a more interesting exploration of when it does well, and when it doesn't against xgboost and also tabm

https://engineering.block.xyz/blog/blocktabbench-evaluating-...

100% agree, bring back LAN messenger, and p2p 'things'. I remember when Opera browser had the 'fridge' which was cool as well (Opera Unite).

Theres definitely ideas here worth exploring for sure.

As an extra data point, I have exactly the same issue on a M1 Max as well.

Business will start accepting bad uptime to be the norm. Following the lead of Github: https://mrshu.github.io/github-statuses/

(Not tested) though apparently it already exists: https://github.com/leejet/stable-diffusion.cpp/wiki/How-to-U...

They gave me 2 queens to start with on my second attempt

Try the pickaxe

Wow, this is a wild ride. I remember coming across this page because the author was from my alma mater and we were pursing the same (undergrad) degree. At the time, we could do a double major in Pure Mathematics and Statistics so long as we completed the coursework requirements, which is probably why that page even exists.

The page is ~15 years old now, and I think it should be read as though its written by a 22 yr old, more reflecting on their recent university education than a guide to how to become a working mathematician.

---

With that note, I would say if someone is eager to engage in mathematics and statistics _at an undergrad level_ (at the time at my university, it was _unusual_ for people to pursue machine learning as a major, and it was in computer science school). I would recommend really focussing on Real Analysis, and the higher statistics courses, try to find the links and the commonality between the proofs and the key ideas. I would also tell myself to not to shy away from martingale theory and link it to measure theory.

Pure mathematics is a weird world. In the moment I hated myself for choosing it in undergrad, it absolutely tanked my grades because of the weird mental state I was in. At the same time when I got to my PhD/research everything starting really started to click. It's immensely difficult to digest and consume all the content in the 12-14 odd weeks that the coursework typically demands.

As someone who isn't particularly well-versed in Bayesian "stuff". Does Bayesian non-parametric methods fall under "uninformative" + "iteration" approach?

I have a feeling I'm just totally barking up the wrong tree, but don't know where my thinking/understanding is just off.

Just an FYI - you can achieve this in 3 layers. It does not need to be deep. https://github.com/charliec443/TreeGrad

Another actuary checking in from Institute of Actuaries Australia - I'm a "qualified" actuary (associate level) who decided to not even try for the fellowship exams and instead change course completely and do a PhD in Machine Learning.

No one outside of the actuarial industry cares or really knows what it means to be an actuary. Having an actuarial background in non-traditional actuarial areas is almost more of a curse than a blessing as people don't really know what to do with you. Furthermore actuaries seem to demand a premium for a cohort that don't have strong enough grounding to do ML research or enough development chops to be an ML engineer. So you end up competing with other people in the data science field...It really is a weird position to be in.

You are definitely correct in saying that in general people view it as simply changing the pronunciations.

Though I would also say that people also appreciate the nuances. There are comedic sketches on this: https://www.youtube.com/watch?v=tK51dlpqpSE (~5.30)

I agree - it looks equivalent to perhaps 1st year mathematics in Australian university or 2nd year 1st semester if I'm being really generous.

This definitely isn't sufficient for machine learning, but it is a start.

h2o.ai doesn't really do data pipelines, though it does appear they are eager to go into this space through their new driverless.ai tool. However this does not appear to be open source.

I have finished the OMSCS program and in some ways I have mixed feelings about it. My background has been primarily in mathematics/statistics and I didn't come from a "tradition CS" educational background.

Did I learn a lot?

I learnt a ridiculous amount. For the time+dollar investment it is amazing. The program is definitely not easy either.

It has been amazing to learn the concepts in ML (Dr. Isbell) and AI (Dr Starner) courses and then a few weeks later think "I think I can actually use these concepts in my workplace".

Why the mixed feelings?

Not all courses had the same quality to it. From the top of my head, AI, ML were probably the best 2 courses. Other well ran courses I would add was computational photography, edutech, introduction to infosec (besides the rote learning...), however some of the other courses I had a relatively negative experience.

The degree does suck up a lot of time and I would say it is the real deal.

Knowing what I know now I can't say 100% that I will "re-do" OMSCS - to be fair on GaTech I'm not sure whether the challenges that I feel above are due to an online program and I personally would be more suited to an in-person program but the experience has definitely been better than Udacity's nanodegree and any MOOC which I have sat.

Overall I would say if you do it for the sake of learning and that alone - OMSCS is worth it. For any other reason please don't do it.

I come from the state of NSW (in Australia), the state with this program, and was also one of the students from selective schools. It is important to note that in Sydney, Australia: it isn't only the public school system which have selective schools, but there exists private schools where academic exams are required to gain entrance (see Sydney Grammar School).

There are a lot of [selective schools; not just one or two, but 46](https://en.wikipedia.org/wiki/List_of_selective_high_schools...)!

You can see in the implementation that some are full selective, and others are partially selective.

Furthermore in primary school there are also [opportunity class](https://en.wikipedia.org/wiki/Opportunity_class) which is also for gifted students. All of these schools are partially selective with around half the spots of high school.

My observation from these schools are that the _location_ of these schools don't matter much. Students will travel from the other side of Sydney to attend any of these schools! For example, there would be a lot of students from the south-west Sydney attending school in North Sydney.

However, segregation based on academic abilities do yield some...results which I have also read about in the US:

* Areas of top selective schools and top OC schools will generally yield higher asian population. * Schools will tend to have a higher asian population

Also (unsurprisingly) these schools have dominated university entrance exams, with the median mark of the top school (James Ruse Agricultural School) typically well within the top 1% of all exam-takers within the state.

Try to do machine learning in Julia; it is difficult. If I want to say fit a gradient boosted tree or SVM its not support in Julia, whilst appears in Python/R libraries. Also, with Spark being more and more popular in the data science landscape, the lack of Julia bindings is also a no for the data scientists I work with (Spark has Python/R bindings).

I currently gave writing a parser combinator in R for a laugh. It might be interesting to someone here. https://github.com/chappers/Ramble

That is unfortunate. I have heard of NetReveal (I used the Detica platform at my previous job) never new that about DataServer though!

I suppose the demand just isn't there. If only more SAS users at least had an interest in programming...

Since SAS is a relatively simple language, why can't someone just write a transcompiler that supports a subset of SAS and move it to R? That way you have the best of both worlds (sort of).

The most difficult thing about that is how you would treat "by" statements (SAS) vs the split-apply-combine (R).

Self-plug: I sort of made a quick hack about a month ago for SAS-Python, I'm sure someone with more programming experience than me could produce something much better (I come from a maths background).

http://nbviewer.ipython.org/gist/chappers/8747253/stan_examp... https://github.com/chappers/Stan

Previous sanctions data analyst here. At my previous bank I worked for they follow a process like this:

1. Screen by name and date of birth (sometimes only year of birth only) for all lists (EU, US, Australia, etc); with some tolerance 2. Manually match against address, sex, etc

Now for some of these lists, depending on the vendor which is used, they may also have information on where the person(s) were last seen (perhaps they're from China and recently visited UK based on public information); not only their country of residence. So this means, if you happen to have a birthday sufficiently close to someone on the list (+-2 yrs) and they happen to be last seen in the country you reside in, then they would have to do the whole manual checking.

Now this seems like a lot of work, but when you consider things like reputation risk of getting caught and the possible fines, banks would rather screw over their own customer (or outright refuse to do certain transactions to certain countries) than doing what is best for their customers.

For the other attributes which are available like sex, it is sometimes used, but often for manual checking. Data is often too messy to have a viable real-time solution for screening. It is often cheaper to get someone from India to do it for you, rather than re-haul a whole database to have it done "properly". This is the unfortunate reality.

You might want to change your spelling of 'receive'