Machine Learning is the new C++

Deep Learning == rebranding of neural nets..its the new hot buzzword, hence the huge amount of downloads. Its high dimensional curve fitting (not a bad thing, if u recognize weaknesses)

 

For Trading, that's a bit of a downer.

But the real money VC allocations tell me that there "may" be more tangible applications since the buzz in the 90s. But then, I didn't know what a VC was in the 90s.

 

I like deep nets, don't get me wrong..just got to be very careful and understand there limitations.

Also, I've heard it is incredible easy to get VC right now, so I'd take it with a grain of salt. Rich old dudes are throwing money at buzzwords.

I did my masters thesis on an aircraft autopilot that could re-learn to fly the jet in the event of damage (lost wing, stuck aileron...etc)...and it worked very well (in simulation). I've also applied them to various other Aerospace applications with success. In my experience, you need to have a solid understanding of the underlying process (domain knowledge). Blindly applying ML algorithms won't cut it.

Good luck

 

Impressive thesis. You had to cross a lot of domains for that.

I am not even entertaining the thought of all those rabbit holes. At the end of the day, I need to make money before I get to retirement age.

 

I'm am Aerospace engineer, and the CS/ML stuff was a hobby, so it worked out

 

Check out this course on neural nets by geoffrey hinton. He covers several types of nets starting with perceptrons, then sigmoid networks, ..., all the way to restricted boltzman machines and self organizing maps.
It's a great and deep introduction to deep nets (pun!) from the man who invented them...

https://gradesfixer.com/blog/neural-networks/

 

Yeah I'm watching that set of videos now. When I have time, which isn't much in this market.

 

http://iknowfirst.com/machine-learning-trading-stock-market-and-chaos

 

pretty interesting read. thnx.

 

IMO we are probably 20-30 years behind what we could have been if we had just called these tools "automated statistical and probabilistic methods" or something like that and divorced the language from these science fiction fantasies.
The most pernicious aspect of this is it creates an artificial benchmark so any actual progress is judged to be unworthy compared to the science fiction fantasy. That has to have real world effects on funding, motivation, time spent on problems, etc.
To say machine learning has no use for trading is like saying a tool box is not good at drilling a hole in the wall to hang a picture. It is missing the most basic concept that the tool box is a container for different tools for different jobs.