Umberto Picchini

@uPicchini

Associate Professor in Mathematical Statistics at University of Gothenburg. Inference for stochastic dynamical systems | Bayes | intractable likelihoods

Vrijeme pridruživanja: prosinac 2012.

Tweetovi

Blokirali ste korisnika/cu @uPicchini

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

  1. Prikvačeni tweet
    16. sij

    PhD studentship with me, on likelihood-free inference (ABC) and the interplay with deep learning. Special (but not exclusive) interest for SDE models. Collaboration with , Andy Golightly and . Up to 5 years in charming Göteborg!

    Prikaži ovu nit
    Poništi
  2. 3. velj

    reminder: we have opened a tenured position for a senior lecturer in statistics and data science. Deadline at the end of the month. This is a department with plenty of success stories. Apply at the link below and explore to know more about us

    Poništi
  3. 30. sij

    for everyone: thanks for sharing your beamer slides on your website. But please recompile using the [handout] option into \documentclass. This will inhibit all the \pause commands. So I won't have to browse 150 pages of an actual 25 slides presentation. Thanks

    Poništi
  4. proslijedio/la je Tweet
    28. sij

    ABC in Grenoble ⛷️, March 19-20: please spread the word by sharing this poster! pdf at: workshop site:

    , , i još njih 2
    Poništi
  5. 26. sij

    "Misra and colleagues found that work days of female proffs were much longer than their male colleagues[...] But the women’s time at work was interrupted and fragmented. Men spent more of their work days in long stretches of uninterrupted time to think"

    Poništi
  6. 24. sij

    A non-standard review of MCMC methods by Christian Robert and Wu Changye. "Non-standard" in that is augmented with questions found on Cross-validated

    Poništi
  7. 22. sij

    6am breakfast while preparing mentally for the first lecture of a new course. By mistake I add white pepper to my muesli instead of cardamom. Great start of the day...

    Poništi
  8. proslijedio/la je Tweet
    17. sij

    New preprint: Joris Bierkens, Sebastiano Grazzi (first paper!), Frank van der Meulen, M.S.: A piecewise deterministic Monte Carlo method for diffusion bridges

    Prikaži ovu nit
    Poništi
  9. 16. sij

    this is part of a joint call for 6 PhD positions distributed across 6 projects on maths for AI. You will be considered for all of the projects but you should mention your specific research interests and indicate which of the projects you are most interested in.

    Prikaži ovu nit
    Poništi
  10. proslijedio/la je Tweet
    16. sij

    Cannot fit all the positive things in 280 characters, so will just say: Go work with Wouter!

    Poništi
  11. 14. sij

    ** yearly reminder for new followers ** subscribe the FREE Bayes Nordics email list, for Bayes-related events & jobs in the (European) Nordic countries. (no need to use gmail to join)

    Poništi
  12. proslijedio/la je Tweet
    14. sij

    We are organizing ABC in Grenoble, March 19-20. With , , Gael Martin, Guillaume Dehaene. Registration and application open at:

    Poništi
  13. 9. sij

    Oh boy, David Blei is an amazing speaker, really. If you have him giving a talk in your area, do not miss it. He talked of variational inference at

    Poništi
  14. proslijedio/la je Tweet
    8. sij

    Also in the ABC session today, presented his clever bootstrap approach to get ABC-MCMC accurate without having to wait until next Christmas. Stratified! As in strata, like a wedding cake.

    Poništi
  15. proslijedio/la je Tweet
    8. sij

    I'm looking for a postdoc who would like to work on ABC, particle MCMC and related topics. Environmental applications. I'll be at the poster session wearing a t-shirt with a basking shark on. Come and talk to me if you are interested.

    Poništi
  16. 8. sij

    Great computational efficiency brought by Prescott's & multifidelity-ABC-SMC compared to standard ABC-SMC, especially when the ABC tolerance is large, which is the nice part as most rejections happen with large tolerance values.

    Poništi
  17. 8. sij

    you ask:"Why continuous time MCMC?" Paul Fearnhead motivates

    Poništi
  18. proslijedio/la je Tweet
    7. sij
    Prikaži ovu nit
    Poništi
  19. 4. sij

    heading to in Gainesville. On Wed 8 Jan I present a work on stratified Monte Carlo and bootstrapping, to alleviate the inefficiency of pseudo-marginal ABC-MCMC with expensive simulators. Joint with . Recently upd preprint at

    Poništi
  20. 2. sij

    I will soon advertise an opening for a PhD on ABC and its interface with deep learning. Plus many other things related to simulations-based inference can happen. A PhD in Sweden is 4-5 years long and salaries are very competitive. If you are interested, please get in touch.

    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:

    ·