moreo
/
QuaPy
1
0
Fork 1
Code Issues Pull Requests Releases Wiki Activity

improved doc

This commit is contained in:
Alejandro Moreo Fernandez 2025-10-09 12:49:08 +02:00
parent d597820a59
commit 1fb8500e87
3 changed files with 12 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
CHANGE_LOG.txt
View File

@ -1,7 +1,7 @@
Change Log 0.2.1 Change Log 0.2.1
----------------- -----------------
... - Improved documentation of confidence regions.
Change Log 0.2.0 Change Log 0.2.0
----------------- -----------------

5
docs/source/manuals/methods.md
View File

@ -604,7 +604,10 @@ estim_prevalence = model.predict(dataset.test.X)
_(New in v0.2.0!)_ Some quantification methods go beyond providing a single point estimate of class prevalence values and also produce confidence regions, which characterize the uncertainty around the point estimate. In QuaPy, two such methods are currently implemented: _(New in v0.2.0!)_ Some quantification methods go beyond providing a single point estimate of class prevalence values and also produce confidence regions, which characterize the uncertainty around the point estimate. In QuaPy, two such methods are currently implemented:
* Aggregative Bootstrap: The Aggregative Bootstrap method extends any aggregative quantifier by generating confidence regions for class prevalence estimates through bootstrapping. Key features of this method include: * Aggregative Bootstrap: The Aggregative Bootstrap method extends any aggregative quantifier by generating confidence regions for class prevalence estimates through bootstrapping. The method is described in the paper [Moreo, A., Salvati, N.
An Efficient Method for Deriving Confidence Intervals in Aggregative Quantification.
Learning to Quantify: Methods and Applications (LQ 2025), co-located at ECML-PKDD 2025.
pp 12-33, Porto (Portugal)](https://lq-2025.github.io/proceedings/CompleteVolume.pdf). Key features of this method include:
* Optimized Computation: The bootstrap is applied to pre-classified instances, significantly speeding up training and inference. * Optimized Computation: The bootstrap is applied to pre-classified instances, significantly speeding up training and inference.
During training, bootstrap repetitions are performed only after training the classifier once. These repetitions are used to train multiple aggregation functions. During training, bootstrap repetitions are performed only after training the classifier once. These repetitions are used to train multiple aggregation functions.

8
quapy/method/confidence.py
View File

@ -339,6 +339,12 @@ class AggregativeBootstrap(WithConfidenceABC, AggregativeQuantifier):
During inference, the bootstrap repetitions are applied to the pre-classified test instances. During inference, the bootstrap repetitions are applied to the pre-classified test instances.
See
`Moreo, A., Salvati, N.
An Efficient Method for Deriving Confidence Intervals in Aggregative Quantification.
Learning to Quantify: Methods and Applications (LQ 2025), co-located at ECML-PKDD 2025.
pp 12-33 <https://lq-2025.github.io/proceedings/CompleteVolume.pdf>`_
:param quantifier: an aggregative quantifier :param quantifier: an aggregative quantifier
:para n_train_samples: int, the number of training resamplings (defaults to 1, set to > 1 to activate a :para n_train_samples: int, the number of training resamplings (defaults to 1, set to > 1 to activate a
model-based bootstrap approach) model-based bootstrap approach)
@ -437,7 +443,7 @@ class AggregativeBootstrap(WithConfidenceABC, AggregativeQuantifier):
class BayesianCC(AggregativeCrispQuantifier, WithConfidenceABC): class BayesianCC(AggregativeCrispQuantifier, WithConfidenceABC):
""" """
`Bayesian quantification <https://arxiv.org/abs/2302.09159>`_ method, `Bayesian quantification <https://arxiv.org/abs/2302.09159>`_ method (by Albert Ziegler and Paweł Czyż),
which is a variant of :class:`ACC` that calculates the posterior probability distribution which is a variant of :class:`ACC` that calculates the posterior probability distribution
over the prevalence vectors, rather than providing a point estimate obtained over the prevalence vectors, rather than providing a point estimate obtained
by matrix inversion. by matrix inversion.