Essays

Predicting human behavior: The next frontiers

See allHide authors and affiliations

Science  03 Feb 2017:
Vol. 355, Issue 6324, pp. 489
DOI: 10.1126/science.aam7032

eLetters is an online forum for ongoing peer review. Submission of eLetters are open to all. eLetters are not edited, proofread, or indexed.  Please read our Terms of Service before submitting your own eLetter.

Compose eLetter

Plain text

  • Plain text
    No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Author Information
First or given name, e.g. 'Peter'.
Your last, or family, name, e.g. 'MacMoody'.
Your email address, e.g. higgs-boson@gmail.com
Your role and/or occupation, e.g. 'Orthopedic Surgeon'.
Your organization or institution (if applicable), e.g. 'Royal Free Hospital'.
Statement of Competing Interests
CAPTCHA

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.

Image CAPTCHA
Enter the characters shown in the image.

Vertical Tabs

  • Inductive and deductive reasoning must be fused for improving prediction accuracy of human behaviors

    Human behaviors are influenced by inductive and deductive reasoning. Current machine learning uses statistical syllogisms. Therefore, the machine learning's conclusion is inherently uncertain. Deductive reasoning is a logical process in which a conclusion is based on the concordance of multiple premises that are generally assumed to be true. In order for artificial intelligence to improve prediction accuracy and behave like human's inference, the conventional machine learning (inductive reasoning) and deductive reasoning must be fused. Prolog (1) and Otter (2) are famous for deductive computer languages. In other words, in order to improve prediction accuracy of human behaviors, machine learning functions (inductive reasoning) must be embedded in deductive computer languages.

    1. https://en.wikipedia.org/wiki/Deductive_language
    2. http://www.mcs.anl.gov/research/projects/AR/otter/

    Competing Interests: None declared.
  • RE: Toward Improved Clinical Decisions
    • Uri Kartoun, Research Staff Member, IBM
    • Other Contributors:
      • Andrew Beam, Research Fellow, Department of Biomedical Informatics, Harvard Medical School

    The essay’s fourth factor regarding the interplay between human behavior and prediction modeling is highly relevant to clinicians and to the decisions that they constantly need to make. In addition to the essay’s authors, many scientists believe that machine learning is revolutionary (1, 2). Many publications describe novel findings of using risk factors to better identify individuals at high risk of developing certain diseases combining highly heterogeneous and noisy unstructured data. Recent advances in software such as code-sharing tools, rapid communication tools, and new algorithms have accelerated health care applications, especially to enhance real-time clinical decision support. Combining the ability to rapidly process the records of hundreds of millions of patients may bring prediction accuracy for diseases and injuries to the level of a genuinely magical crystal ball. Such abilities were not sufficiently available to us a decade ago.

    Our paper appeared Feb. 9 in Scientific Reports (3) illuminates an intriguing pattern in how physicians provide care. By analyzing the electronic medical records of more than 1,000 patients experiencing a sleep disorder, we demonstrated that a physician’s decision on which drug to prescribe does not primarily rely on a patient’s condition. Instead, a physician’s habit is the main driver to prescribing a certain drug over another. Applying natural language processing techniques and machine learning algorithms played a crucial r...

    Show More
    Competing Interests: None declared.