Machine Learning in Finance -- Coursera

Discussion in 'Automated Trading' started by ajensen, Jun 1, 2018.

  1. ajensen

    ajensen

    The NYU Tandon School of Engineering has created a Machine Learning and Reinforcement Learning in Finance Specialization with four courses on Coursera:

    https://www.coursera.org/specializations/machine-learning-reinforcement-finance
    (1) Guided Tour of Machine Learning in Finance
    (2) Fundamentals of Machine Learning in Finance
    (3) Reinforcement Learning in Finance
    (4) Overview of Advanced Methods of Reinforcement Learning in Finance

    I have enrolled in the second course. There is a one week free trial, after which it costs $39/month.

     
    dtrader98 likes this.
  2. Maverick74

    Maverick74

    I would highly reccomend you checking out Quantstart. Same cost, but a lot more value.

    https://www.quantstart.com/
     
  3. trader99

    trader99

    Both of these are great resources. But making money in the market isn't as easy as feeding data into machine learning algorithms and alpha comes out of it. There's still a long hard road ahead...
     
    userque and vikasd like this.
  4. ajensen

    ajensen

    That's true. For some markets that interest me, and which I have studied, I have created some indicators that predict daily returns in a multiple linear regression. I want to see if the same indicators can predict better when using methods such as SVM, random forests, and neural networks.
     
    trader99 likes this.
  5. ph1l

    ph1l

  6. Hi Mav,
    Would you elaborate a bit your preference for the Quantstart course over the Coursera offering. Was not familiar with Quantsart ... looks interesting.
     
  7. Maverick74

    Maverick74

    Quantstart is more of a community like ET. So you can exchange ideas with others, learn to code, read about how these algos actually work in trading. Coursera is great and I love MOOCs in general, but it's just a course, and really a theoretical look at the backbone of machine learning. The question is, how do you put it to use in real life. That's where most traders need help.
     
  8. Waste of time. Too complex to generate results. In trading the simplest model wins.
     
    gkishot likes this.
  9. Little of the academic staff in quantstart work in practice. It's good stuff for academic research but not for practical trading. Look for people like Cesar Alvarez for practical trading tips and methods. Academic Phds overwhelm people with terminology and useless obfuscation. By the time you get around to digest all that useless terminology you are probably done trading.
     
  10. Maverick74

    Maverick74

    I'll pass your comments onto to Jim Simons and Ed Thorp.
     
    #10     Jun 24, 2018
    sle and tonysoprano like this.