Serving the Quantitative Finance Community
Machine learning for Quant
Hi,would the knowledge, understanding and experience in machine learning/statistical learning(probably its called by many other names) be useful in career as Quant? may be not directly, but indirectly?I mean, for example the mathematics involved in supervised learning(svms, glm, bayesian networks etc), unsupervised learning(SVD, EM algorithms etc) could be used anywhere ?Thanks.
Machine learning for Quant
quants without even basic statistical modelling skills are abound, especially on the sell side. Any knowledge and a bit of practical experience in ML would give you an unfair advantage and is, therefore, frowned upon...
Machine learning for Quant
QuoteOriginally posted by: yuryrquants without even basic statistical modelling skills are abound, especially on the sell side. Any knowledge and a bit of practical experience in ML would give you an unfair advantage and is, therefore, frowned upon...Frowned upon? does that mean Interviewers might not like you?
Machine learning for Quant
QuoteOriginally posted by: ShaikQuoteOriginally posted by: yuryrquants without even basic statistical modelling skills are abound, especially on the sell side. Any knowledge and a bit of practical experience in ML would give you an unfair advantage and is, therefore, frowned upon...Frowned upon? does that mean Interviewers might not like you?You are a very smart guy. Just in case you mention SVM or similar topics be sure that you know linear regression really-really well 
Last edited by tu160 on April 2nd, 2012, 10:00 pm, edited 1 time in total.
-
- ArthurDent
- Posts: 5
- Joined:
Machine learning for Quant
Go to buy side - hedge funds.Avoid sell side - investment banks.Find a good head hunter.Good luck.
-
- Anomanderis
- Posts: 0
- Joined:
Machine learning for Quant
C'mon guys, can you at least advise why ML is frowned upon?
Machine learning for Quant
Quotea bit of practical experience in ML would give you an unfair advantage and is, therefore, frowned upon...- it was a joke...clearly you want to have an unfair advantage.but a) ML is used for "quant trading" - mainly hedge funds as AD said.b) majority(?) of quants are in investment banks - doing derivatives, where you act as a marketmaker/bookmaker rather than betting yourself. so ML not useful.c) as TU says, you should understand basic statistics first (and be able to demonstrate that) interview with quant prop trader
Last edited by spv205 on April 2nd, 2012, 10:00 pm, edited 1 time in total.
-
- igemonster
- Posts: 0
- Joined:
Machine learning for Quant
QuoteOriginally posted by: AnomanderisC'mon guys, can you at least advise why ML is frowned upon?I think he was only kidding..
Machine learning for Quant
to add a bit more color, mentioning ML and statistics in general might attract some critisism upfront on interview in IB because: a) many people indeed mention all sorts of ANNs, SVMs etc.. on their CVs but unable to solve simple problems, let alone make ANNs work for a particular application, b) interviewers sometime do not even understand the language, some (but not all!) "big" quants in banks will never admit that they cannot fit any model beyond simple linear regression reliably