HNHacker News
TopNewBestAskShowJobs

anjc

1,414 karma · joined October 24, 2011

submissionscomments
anjc··on Any claim without a URI should be treated as suspicious
Even source data itself is not free from contexts and biases. Just treat everything with suspicion. And treat people with extra suspicion if they always provide a URL, especially if the URL is to somebody's interpretation of an interpretation etc.
anjc··on The AirPods Pro “Rattlegate”
The XPS 15 throttles horribly at reasonable temperatures. Many people re-seat the heatsink on the CPU and GPU and it fixes this issue.
anjc··on Ten Lessons I Wish I Had Been Taught (1997) [pdf]
> they aggressively re-publish their existing results. Very low read- and citation counts are common implications.

I think it's generally the opposite to a great degree. Some people spend their PhD and postdoc doing one large piece of work, overly concerned about its rigour/correctness/validity, eventually publishing just one paper that they're happy with (perhaps a summary of their thesis). Others publish smaller pieces of work frequently, at conferences, workshops etc. In this time the latter will have built up a public academic profile, while the former may be essentially anonymous to the field. This will be reflected in greatly more citations and reads for the latter.

You could argue that there's value to the former approach, but if you can't disseminate your work effectively - i.e. nobody knows that you or your work exist - then it's useless (imo).

anjc··on MS will ban Forza players who add the confederate flag to their digital cars
How can seceded states commit treason against a union that they aren't part of?

> resulted in the mass murder of hundreds of thousands

In other words, a war

anjc··on Tell candidates what to expect from your job interviews
> We used to apply a ‘Sunday Test’: “If this person were alone at the office on a Sunday, would that make us want to come in just to hang out with them?”. However, this was the wrong phrasing for the idea we really cared about (we don’t encourage people to work Sundays, and we want people who would make excellent coworkers rather than be our personal friends). Instead, we now ask: “Is this someone we’d actively seek to work with”

I'd love if this requirement were explicable and quantifiable. Two people can each be nice and kind, but not get along for whatever reason. When a hiring process can include face to face meetings with 10+ people, the chances of a suitable candidate being excluded for an arbitrary 'personal' reason seem high.

What's the point in many stringent evaluation criteria if, at the end of it, all interviewers then have to agree that they can be besties with a random person. It's too fuzzy.

anjc··on Does Tweeting Improve Citations? One-Year Results from a Randomized Trial
> This means that many papers citation list is not a result of a thorough literature search

Related work sections aren't typically thorough, that's what lit surveys/reviews are for. It's also not a new phenomenon that highly cited work gets cited because it's highly cited, i.e. work is cited simply because it's more readily noticed. It's furthermore common for the same research to arise independently in different fields, with different terminology, i.e. bodies of research are somehow not found. So, no need to be scared, it's nothing new.

anjc··on GitHub to replace “master” with alternative term to avoid slavery references
It seems awkward to have a verb as part of the noun, but as I write this I'm thinking of many.
anjc··on GitHub to replace “master” with alternative term to avoid slavery references
I don't understand the racist connotations given that there are no "slave" branches. Every similar mention of "master" can't change. What words can better describe a master boot record, for example?
anjc··on GitHub to replace “master” with alternative term to avoid slavery references
Why does it make more sense? Blacklist is used also as a verb, and I don't think "denylist" could be used as such. You could say "add to the denylist" but I don't think this is sensible.
anjc··on Woodworking for Engineers
If you want a functioning shop, and you were limited to four initial options:

1) prohibitively expensive power tools

2) prohibitively expensive hand tools

3) cheap hand tools

4) cheap power tools

...I'd go for cheap power tools every time. You could get more tools than you'll ever need for $1-2k buying cheap but usable brands. They probably wouldn't stand up to more than DIY use, but they will work, and you'll learn what items you really value. For example, I now know that I use my cheap table saw enough that I could justify a more expensive one, however I'll never buy an expensive scroll saw.

I think it's bad advice to start with good hand tools unless you have a high tolerance for manual labour or are apprenticing under someone in a fine woodworking area (e.g. luthiery).

anjc··on Potential organized fraud in ACM/IEEE computer architecture conferences
It is of course good to have the support of your supervisor, but if the relationship breaks down then you shouldn't feel that your graduation is at risk because of this. I think most universities will have a policy on changing supervisors that favours the student? Whether or not this affects your future career I don't know.
anjc··on Potential organized fraud in ACM/IEEE computer architecture conferences
Well thanks for that, I'm confused about the relevance of an 'advisor' to a paper from the perspective of the PC or publisher, or how they would even determine who an advisor is. Do IEEE/ACM have rules for this or is this part of the alleged fraud?
anjc··on Potential organized fraud in ACM/IEEE computer architecture conferences
It isn't up to his mentor to withdraw a paper for which he isn't the first author. It doesn't even require a retraction as it wouldn't be published yet, albeit, withdrawing is still frowned upon
anjc··on Guitar Decomposed: 5. Mutating the Third
> "Step" refers to an interval gap of "half step" or "whole step"

Right, but it does not mean 'half step'. It means one or other depending on the harmonic context, not the language context as said above. In the case in the article, a "step" down from the third is the second, not the minor third as the article says.

> I would be quite open to a definition where '1 step' is always '1 half step'

I don't understand this, sorry. Why not just use semitone?

> Since both 'semitone' and 'half step' imply a less useful fundamental value of 2 half steps

Western music isn't fundamentally chromatic either, so talking in semitones isn't particularly useful.

As you imply with reference to wholetones and dissonance, the words tone/semitone also give useful melodic information that we've all already learned. E.g. raising the 5th a tone is consonant, raising the major 3rd a tone is dissonant. Of course you could teach students that not all steps/two steps are the same, depending on scale degree, but then you're back to square one, just with different words for semitone/tone.

> Your entire hypothetical dialogue is confusing as well, since your hypothetical person isn't using correct words either.

Yeah it's meant to be confusing, my point is that overloading terms gets confusing quickly, at least for me.

anjc··on In NY major crime complaints fell when cops stopped ‘proactive policing’ (2017)
> car breakins are not infrequent in the urban US. Ours was probably particularly frequent because it was being targeted by some sort of organized setup, but it is not uncommon.

> Generally sounds like you'll probably want to stay in suburbia.

So the poster you replied to was correct to say that they can just move to cities where it isn't accepted that cars are frequently broken into. They don't have to stay safe in suburbia after all.

anjc··on In NY major crime complaints fell when cops stopped ‘proactive policing’ (2017)
> I live in a very nice area of DC, cars would get rifled through every ~2 months or so

> Generally sounds like you'll probably want to stay in suburbia.

It rather sounds like you have a strange perspective on urban living. I can't speak for the US, but it isn't normal in any city in Europe for your car to be rifled through every 2 months in a nice area of a city. It isn't even normal for bad areas of cities. If it happens once in a person's life it would be bad luck.

anjc··on Guitar Decomposed: 5. Mutating the Third
> Depends on your language use

'Step' has a specific meaning in music and it doesn't mean 'semitone'. Why not just use semitone seeing as that precisely describes what's intended? Conflating meanings seems like it could be confusing. "The tritone of this note is 6 steps up" "Gotcha, so the tritone of C is B" "No no, by step I currently mean tones, not semitones" "Ah, so 3 scale tones, ok" "No just 3 tones, or 6 steps" "Ok, I'll count 6 steps, 1,2,3,4..."...just use the correct word.

anjc··on Sid Meier’s Civilization VI free until May 28th
> I played a lot of VI and stuck with it for a few years

> It was a fun game and I got my value out of it

> you might want to pass on this [game that I played for years which is temporarily free]

?

anjc··on Spotify signs ‘The Joe Rogan Experience’ to an exclusive multi-year deal
It work the opposite way too, wherein platforms lose users due to undesired changes. YouTube's banning/striking/demonitizatation of high profile channels, something that Joe often says he's afraid of, could be seen as an undesirable platform change.
anjc··on Spotify signs ‘The Joe Rogan Experience’ to an exclusive multi-year deal
It doesn't take much listening to JRE or similar shows to realise that they're in constant fear of strikes and demonetization, even when talking about mundane topics. Joe's no Alex Jones or David Icke, and he's still afraid of getting more strikes on the channel. If this move means he can talk comfortably then good for him. It's not as if listeners lose any rights in the process.

Would this mean he could play artist's music that Spotify are allowed to stream? Or would other agreements be needed for this?

anjc··on So much of academia is about connections and reputation laundering
>What’s striking is that the professor and A/Chairman doesn’t know that he doesn’t know. I’m struck by his ignorance of his ignorance, his willingness to think that he knows what he’s talking about when he doesn’t.

Does this make sense to anybody? The author is "struck" that somebody is "willing" to talk about something they believe they understand? The only annoying thing about academia is that it seems to attract smug, selfish people like this.

anjc··on So much of academia is about connections and reputation laundering
> your citation count, number of papers, awards etc quickly becomes insignificant compared to those who are

There are few metrics with which academia can judge how effective a researcher is. If you're not interested in getting as many papers/citations/awards in whatever way you can then you may be in the wrong game.

anjc··on Applying to PhD Programs in Computer Science (2014) [pdf]
> I'm from Europe so I wouldn't be in debt or anything

Do a Masters and it will help with all of your questions

anjc··on Applying to PhD Programs in Computer Science (2014) [pdf]
> Go into grad school and pursue a PHD because you've got a burning desire to turn one of your good ideas into a breakthrough.

I don't think many supervisors will let you do this. 1) you'll need to find someone who has funding and already works in the area of your good idea 2) people overestimate how much they can do during a PhD 3) you probably aren't aware of the field coming out of undergrad 4) not a lot of theses will ever be read outside of the examiners, let alone be a breakthrough.

I'd say do a PhD if you feel like it, you can work for years on mundane tasks that likely wont culminate in a breakthrough, aren't prone to depression/stress, and don't care about money.

anjc··on A List of Colleges’ Plans for Reopening in the Fall
Oh I checked my uni's calendar and you're right, there's a brief notice that deferral isn't guaranteed. In practice I've never seen it denied, but these times are different.
anjc··on Google Research Football
Are there any regulations against transmitting data to wearables during play? I only see a regulation saying that data from wearables can't be used during a game. Is there anything preventing players from wearing e.g. a device that vibrates when they're not optimally positioned, according to camera tracking
anjc··on Lithuanian capital to be turned into vast open-air cafe
> Car centered cities are awful for human beings

Do you just mean that a car centred city is more noisy and dirty than an economically-equivalent city that has fewer cars? They've been amazing for human beings if you take a broader/longer perspective.

anjc··on Using Reinforcement Learning in the Algorithmic Trading Problem
> They also don't compare to baselines

This is certainly worth criticising

> they can still overfit to test data by running a lot of experiments with different hyperparameters, architectures and parts of the data, and only report what has worked.

but this is a different accusation from accidentally overfitting or leaking, i.e. it would mean that they're dishonest and cherrypicked their data in such a way that it hides overfitting and leakage. This criticism can be levelled at every ML paper, but in this case they detail their architecture, provide the code, and provide a Jupyter notebook to let people try it themselves.

> just assume they can trade at whatever price the data tells them. It's completely unrealistic.

I think that this is a fair assumption for highly liquid markets and relatively small trades, and if it's a fair assumption then all of your criticisms (slippage etc) don't apply to the extent that they'll break the approach. Also, if the approach works then trade size (fees aside) and being frontrun also wont apply because presumably large HFT firms can use it.

Overall I think your criticisms are valid, but imo they don't invalidate a promising approach, they're just the next thing to test.

anjc··on Using Reinforcement Learning in the Algorithmic Trading Problem
How do you propose that anyone evaluates their algorithm on the future?
anjc··on Using Reinforcement Learning in the Algorithmic Trading Problem
Many people are criticising their backtest but I don't understand why. Their test data is sequential to their training data, considers time, and doesn't overlap. They can't overfit to their test data. In any other area of ML this would be an acceptable scheme, why is this unacceptable here?
← PreviousPage 5 of 33Next →