1. Guest, click here and read the 'News in Brief' post for w/c 8th July.
    Dismiss Notice

'AI' extracts new knowledge from old science papers

Discussion in 'Health News and Research unrelated to ME/CFS' started by Marco, Jul 14, 2019 at 8:57 AM.

  1. Marco

    Marco Senior Member (Voting Rights)

    Messages:
    202
    Likes Received:
    931
    In the domain of thermoelectric materials only for the time being but this machine learning program was able to 'predict' later findings from older papers :

    https://www.zerohedge.com/news/2019...ntific-papers-makes-discoveries-missed-humans

    But in principle the approach could be extended to other areas :

    I doubt there is enough good science around ME/CFS for this approach to be directly useful but advances in understanding the immune system; metabolimics; the central nervous system etc may be relevant.

    Original paper abstract : https://www.nature.com/articles/s41586-019-1335-8
     
  2. Andy

    Andy Committee Member & Outreach

    Messages:
    6,689
    Likes Received:
    49,798
    Location:
    Hampshire, UK
    I'm sure @mariovitali would support the message from this paper.
     
  3. mariovitali

    mariovitali Senior Member (Voting Rights)

    Messages:
    258
    Likes Received:
    1,365
    I will say it depends @Andy.

    We do not want to be employing AI methods for the sake of it or because it is a buzzword. We need tangible results and thus prove that this technology outperforms other approaches.


    I do agree with the paper in the sense that this technology can be used for medical research. However we have to make sure that the data is of good or sufficient quality and also to make "reality checks" very often to identify whether the software works as expected or not.

    I was really happy to see that AI and Network Analysis was mentioned in the NIH conference but also in SolveCFS where Machine Learning and AI are discussed as technologies that "could help move the needle". It remains to be seen if this will be indeed the case.
     
  4. Forbin

    Forbin Senior Member (Voting Rights)

    Messages:
    728
    Likes Received:
    5,653

    The AI's going to be pretty frustrated when it gets to the PACE trial.
     
  5. Barry

    Barry Senior Member (Voting Rights)

    Messages:
    4,379
    Likes Received:
    29,218
    Kind'a puts a new slant on "Artificial Intelligence" don' it ;).
     
  6. rvallee

    rvallee Senior Member (Voting Rights)

    Messages:
    1,825
    Likes Received:
    14,776
    Location:
    Canada
    E X C E P T I O N F O U N D

    W E H A V E L O S T I N T E L L I G E N C E
    [​IMG]
     
  7. NelliePledge

    NelliePledge Senior Member (Voting Rights)

    Messages:
    3,000
    Likes Received:
    21,587
    Location:
    UK East Midlands
    PACE isn’t appropriate for AI it is appropriate for GT. Genuine tripe. Sorry couldn’t resist. :sneaky:
     

Share This Page