do u even machine learning bruh

thats the question you should ask yourself if you're studying physics, biology, engineering, maths - anything scientific

this is where the money is going to be

you heard it hear first.

Reply 1

Ayygeee

OP

5

maybe. But there are plenty of graduates with so-called strong quantitative degrees that have no idea

The training you need to tweak network architecture to fit the subtleties of applied ML problems isn't that commonly available unless you pick the very ML-orientated courses as part of your quantitative degree or take a masters in ML.

Reply 2

artful_lounger

Universities Forum Helper

21

Some people pursuing STEM subjects aren't interested in inserting their heads into the CEO of Google/Facebook/etc's arses to the end of of rampantly abusing people's personal data for the corporate bottom line...no accounting for ethics I suppose.

Also lets also stop using the euphemistic "machine learning" to obscure your specific aim, which is pretty clear from your phrasing - it's data mining. As tawdry an occupation as it sounds...

Reply 3

Ayygeee

OP

5

Original post by artful_lounger

Some people pursuing STEM subjects aren't interested in inserting their heads into the CEO of Google/Facebook/etc's arses to the end of of rampantly abusing people's personal data for the corporate bottom line...no accounting for ethics I suppose.

Also lets also stop using the euphemistic "machine learning" to obscure your specific aim, which is pretty clear from your phrasing - it's data mining. As tawdry an occupation as it sounds...

No absolutely not. You have a limited appreciation of what machine learning can accomplish.

Have a read: https://www.nature.com/articles/nature21056

(edited 5 years ago)

Reply 4

gjd800

20

Original post by Ayygeee

No absolutely not. You have a limited appreciate of what machine learning can accomplish.

Have a read: https://www.nature.com/articles/nature21056

That is incredibly cool.

Reply 5

Ayygeee

OP

5

Are you saying that if you understand regression ... you have the skills to convert a published architecture to something that better fits your particular problem eg. taking a 2D conv net into a 3D architecture, changing the regularisation, combining a conv net with a RNN to apply to short video clips... all kinds of thing.

I agree with you ... to some extent ... regarding the MSc. The problem is that for a lot of bad universities, this doesn't hold true. The only way these people can get some credibility is to do something like the 1 year UCL MSc.

Reply 6

PossiblyNotGod

20

xD this is entertaining

Reply 7

artful_lounger

Universities Forum Helper

21

Original post by Ayygeee

Are you saying that if you understand regression ... you have the skills to convert a published architecture to something that better fits your particular problem eg. taking a 2D conv net into a 3D architecture, changing the regularisation, combining a conv net with a RNN to apply to short video clips... all kinds of thing.

I agree with you ... to some extent ... regarding the MSc. The problem is that for a lot of bad universities, this doesn't hold true. The only way these people can get some credibility is to do something like the 1 year UCL MSc.

Pray tell, how much is UCL paying you to advertise their course...?

Reply 8

Ayygeee

OP

5

Original post by artful_lounger

Pray tell, how much is UCL paying you to advertise their course...?

Its just an example of a good course. I think there are plenty of good ones.

Reply 9

UWS

21

Genius is back.

It's been a while.

bump

OP

Yeah but for quite narrow use cases and within the Apple ecosystem. For the majority of companies looking to be innovative, you need way more expertise than you seem to think.

I would say what you're talking about is someone who drives a car, I am talking about someone who designs a car.

(edited 5 years ago)

brah

this is now a meme page:

OP

lol i loved how excited you are by this

Tell me what you do, im hiring

Reply 18

Profesh

20

Original post by artful_lounger

Also lets also stop using the euphemistic "machine learning" to obscure your specific aim, which is pretty clear from your phrasing - it's data mining. As tawdry an occupation as it sounds...