Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Using a financial training criterion rather than a prediction criterion

Y Bengio1

  • 1Department of IRO Université de Montréal, Qc, Canada. bengioy@iro.umontreal.ca

International Journal of Neural Systems
|August 1, 1997
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Adaptive importance sampling to accelerate training of a neural probabilistic language model.

IEEE transactions on neural networks·2008
Same author

Taking on the curse of dimensionality in joint distributions using neural networks.

IEEE transactions on neural networks·2008
Same author

Cost functions and model combination for VaR-based asset allocation using neural networks.

IEEE transactions on neural networks·2008
Same author

Experiments on the application of IOHMMs to model financial returns series.

IEEE transactions on neural networks·2008
Same author

Bias learning, knowledge sharing.

IEEE transactions on neural networks·2008
Same author

Locally linear embedding for dimensionality reduction in QSAR.

Journal of computer-aided molecular design·2005

This study shows that directly optimizing financial gains and losses improves decision-making in noisy financial time series. This approach outperforms traditional prediction criteria for trading strategies.

Area of Science:

  • Computational finance
  • Machine learning applications
  • Financial econometrics

Background:

  • Traditional financial time series analysis often relies on prediction criteria like minimizing squared error.
  • These methods may not directly align with maximizing actual financial performance, especially with noisy data.
  • Direct optimization of financial objectives offers a potential improvement.

Purpose of the Study:

  • To investigate the efficacy of directly optimizing financial criteria (gains and losses) versus traditional prediction criteria for financial time series.
  • To enhance decision-making models in finance using machine learning.
  • To evaluate the performance in a real-world portfolio selection scenario.

Main Methods:

  • Development of a learning algorithm trained to directly maximize financial gains and losses, including transaction costs.

Related Experiment Videos

  • Comparison of this direct optimization approach against traditional prediction-based training methods.
  • Experimental validation using a portfolio selection task on 35 Canadian stocks.

    • Main Results:

    • Models trained to directly maximize financial criteria yielded superior results compared to traditional prediction-based models.
    • The direct optimization approach proved more effective in handling noisy financial time series data.
    • Demonstrated improved performance in portfolio selection tasks.

    Conclusions:

    • Directly optimizing financial performance metrics is a more effective strategy for decision-making in noisy financial time series.
    • Machine learning models can be successfully adapted to prioritize financial outcomes over mere prediction accuracy.
    • The findings have significant implications for algorithmic trading and portfolio management.