Discussing the article: "Exploring Conformal Forecasting of Financial Time Series"
MetaQuotes:
Very interesting and informative article. I will reread it again and try to implement the approach in my MT5 terminal. Thanks to the author.
Published article Exploring conformal forecasting of financial time series:
Author: Maxim Dmitrievsky
I have used conformal mappings in hydro-aerodynamics, but to apply to financial time series is interesting. I don't agree with the author on everything, but respect anyway.
Hi, I think you forgot to attach the fixing_lib module. The module is being imported in the file mapie_causal.py
Great work! Thank you very much for the contribution. I made some adjustments and it works perfectly.
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Check out the new article: Exploring Conformal Forecasting of Financial Time Series.
MAPIE (Model agnostic prediction interval estimator) is an open-source Python library designed for quantifying uncertainty and managing risk in machine learning models. It allows computing prediction intervals for regression problems, as well as prediction sets for classification and time series. This uncertainty assessment is performed based on a special "calibration set" of data.
One of the key advantages of MAPIE is its model-agnostic nature, which means the library can be used with any model that is compatible with the scikit-learn API, including models developed using TensorFlow or PyTorch, through appropriate wrappers. This property greatly simplifies integration into existing analytical pipelines, as traders often use a variety of machine learning models, from traditional statistical approaches to complex neural networks, depending on the specific asset class or trading strategy. The ability to seamlessly use proven models to incorporate uncertainty quantification significantly reduces implementation costs and accelerates adoption, which is particularly valuable in a dynamic financial environment.
The library is part of the scikit-learn-contrib ecosystem and builds on conformal forecast and distribution-free inference. It implements peer-reviewed algorithms that are model- and use-case-independent and have theoretical guarantees with minimal assumptions about the data and model. Beyond standard classification, MAPIE is also capable of risk control for more complex tasks such as multi-class classification and image segmentation in computer vision by providing probabilistic guarantees on metrics such as recall and precision.
Author: dmitrievsky