Alexandr Proskurin

@proskurinalex

Quantitative Researcher/Algorithmic Trader, co-author of mlfinlab package

Kyiv
Vrijeme pridruživanja: ožujak 2012.

Tweetovi

Blokirali ste korisnika/cu @proskurinalex

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @proskurinalex

  1. Prikvačeni tweet
    3. velj

    0.5.0 release is ready! Key features added: - Structural breaks tests. - Microstructural and entropy features. - Constant tick size imbalance/run bars. - Time bars. - Optimal Number of Clusters algorithm (ONC). The whole list of changes can be found in release notes.

    Poništi
  2. 5. velj

    Recently posted great lecture slides on variable codependence. Our research team implemented all of them and included into release. We decided to test metrics on various non-linear variable dependencies and results look great!

    Poništi
  3. proslijedio/la je Tweet
    14. sij

    Wrote a blog post on the Hierarchical Risk Parity (HRP) algorithm for Machine Learning Asset Allocation. I try to explain the intuition and math behind the algorithm and do a performance analysis with other allocation algorithms. Post -

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Poništi
  4. 6. pro 2019.
    Poništi
  5. 20. stu 2019.

    Excited to give a lecture with at the London School of Economics on the topic of the financial machine learning, its applications, key techniques and how is used in a full-scale machine learning quantitative research!

    Poništi
  6. 18. stu 2019.

    Well-known PIN/VPIN is just one of many microstructural features that can be used in HFT algorithms. Kyle, Amihud, Hasbrouck lambdas, tick rule entropies may contain useful information for algorithmic traders. All of them will be added in the next release!

    Poništi
  7. 12. stu 2019.

    9 Pull Requests pending in repo, lots of code to review this month. We are on the way to finish all the material described in Advances in Financial Machine Learning book!

    Poništi
  8. 5. stu 2019.

    PR for Structural Breaks chapter is ready! Managed to speed-up tests (SADF, Chow-Type, Chu-Stinchcombe-White) via multiprocessing, however as said, SADF test computation time is still huge! Any thoughts on the algorithm complexity structure improvements?

    Poništi
  9. 3. stu 2019.

    Getting ready for a new mlfinlab release. Introducing constant tick size imbalance/run bars for those facing the issue of imbalance bars converging to ticks. Stay tuned!

    Poništi
  10. 10. ruj 2019.

    For those of you looking how to solve the problem of label dependency in financial machine learning we have posted an article which describes Sequential Bootstrapping algorithm and how it can be applied to solve the problem of label concurrency:

    Poništi
  11. 4. ruj 2019.

    Really excited about the latest release of mlfinlab () based on book. We have managed to get a remarkable performance increase x15000 times in Sequential Bootstrapping implementation! Thanks to the team for hard work and passion!

    Poništi
  12. proslijedio/la je Tweet
    3. ruj 2019.

    The latest version of mlfinlab () has been released. Time to: "pip install mlfinlab" This new release is our biggest to date and includes code for: 1. Sequentially Bootstrapped Ensembles (Regression & Classification) ( )

    Prikaži ovu nit
    Poništi
  13. proslijedio/la je Tweet
    27. kol 2019.

    The Single Futures Roll []

    Poništi
  14. proslijedio/la je Tweet
    16. lip 2012.

    Росія сенсаційно вилетіла з чемпіонату Європи

    Poništi
  15. 24. tra 2012.

    Тактичность и корректность украинских подъездов.

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·