Mathophile

PhD student in Applied Mathematics with a focus on machine learning. Also giant nerd for D&D, comics, and a political junkie. Speak civil and think critical.

[Mathophile @mathophil2_718]

@MuffugginGoblin @pseudocia @Gasparatus Yeah scale matters. If you're already gaming and you only care about your room, then yeah you probably don't need a space heater, your computer is the space heater. Plus, your gaining dual use. But a whole warehouse as a public shelter, you'd definitely want better.

[Mathophile @mathophil2_718]

@MuffugginGoblin @pseudocia @Gasparatus For space heaters sure. But that's a fairly weak form of heating. You'd be served better comparing to furnaces. I'm not surprised by a space heater performing comparable to a computer. Tgey just draw power through the coils use the resistance to create heat.

[Mathophile @mathophil2_718]

@pseudocia @Gasparatus Energy loss is any energy that leaves the system. That loss can take different forms than heat. For example, in a car crash, bumpers that break and make large crunching sounds are actually dispersing some of the kinetic motion into sound. Thus sound is the energy loss.

[Mathophile @mathophil2_718]

@specksack @pseudocia @Gasparatus What don't you get about my statement? Computers utilize electricity to change dipole moments, not to generate great. Heat is energy loss for a computer. Heaters on the other hand are designed to efficiently generate great and disperse it.

[Mathophile @mathophil2_718]

@kareem_carr Honestly, I think that's a unfair assessment. I can see why you might come to that conclusion of you're judging it based on news. But you'll suffer from sample bias, as people in public aren't a representative example of the population.

[Mathophile @mathophil2_718]

@gwisk20 Why feel shame? Do you know how many mathematicians can't do simple calculations off the cusp? Use a calculator, you've got better things to do than crunch a number!

[Mathophile @mathophil2_718]

@pseudocia @Gasparatus While true, that energy conversion is an inefficient conversion of energy types. The heat generated is an energy loss of the computer. But it takes considerably more energy for a computer to raise the temperature of a room than a heater.

[Mathophile @mathophil2_718]

@pseudocia This is also similar to the concept of using machine learning to work on climate issues. The supercomputers used to train ml algorithms consume a lot of energy.

[Mathophile @mathophil2_718]

@Gasparatus @pseudocia Except that you could easily keep a shelter warm via electric or gas heating at a cheaper price and with less carbon footprint.

[Mathophile @mathophil2_718]

@GarySeinfeld5 @vbarot87 @pseudocia Can I get the link to this paper?

[Mathophile @mathophil2_718]

@MarcosPolanco @Light_bound @MaxCRoser Keep in mind with that table, it's for US households. There also exists a large wealth gap between US and other nations of the world. So even that distribution is heavy tailed.

[Mathophile @mathophil2_718]

@FuTB12 @AlexPadilla4CA You need to compare wages vs CPI to evaluate purchasing power. Avg wages have been increasing at a similar to slightly higher rate than CPI, but median wages are stagnant while CPI is increasing. So lower earners are losing purchase power.

[Mathophile @mathophil2_718]

@AnalysisFact Proof of the claim is trivial, and left as an exercise to the reader.

[Mathophile @mathophil2_718]

@_Michael_Selasi @WillSpagnoli @svpino Coursera and MIT have some great online work. My suggested math subjects for ML are: probability theory (not stats), linear algebra, numerical analysis (basic, linear alg, and optimization). Optimization is the goal, and then realize that ML is a form of nonlinear optimization.

[Mathophile @mathophil2_718]

@svpino Programming takes up the majority of my time, but it's the least meaningful aspect of my time. Coding is just the to you use so you can analyze. You need to learn sme of some kind to make any useful analysis beyond, I got better results.

[Mathophile @mathophil2_718]

@phdepresso @AcademicChatter Unfortunately, you are in the rough. Likely your department won't support you and you'll have an uphill battle with hr. Something you have to consider is the value-stress trade-off of pushing the issue compared to finding a new advisor. It shouldn't be this way, but it is.

[Mathophile @mathophil2_718]

@kareem_carr Even 1 day old coffee feels like it's mugging me.

[Mathophile @mathophil2_718]

@MarissaKawehi So they were actually giving productive criticism, nice.

[Mathophile @mathophil2_718]

@shoe0nhead I'll go ahead and get rid of genetics, life situations, and prevent the actions of others. That'll cover me.

[Mathophile @mathophil2_718]

@ravenscimaven How important are mathematical models to your research? Do you find your research guided by models, or the other way around?