Skip to main content

An implementation of Exact Soft Confidence-Weighted Learning

Project description

The explanation of the algorithm

This is an online supervised learning algorithm.
This learning method enjoys all the four salient properties:
  • Large margin training

  • Confidence weighting

  • Capability to handle non-separable data

  • Adaptive margin

The paper is here.

There are 2 kinds of implementations presented in the paper, which served as

scw.SCW1(C, ETA)
scw.SCW2(C, ETA)

in the code. C and ETA are hyperparameters.

Usage

from scw import SCW1, SCW2

scw = SCW1(C=1.0, ETA=1.0)
weights, covariance = scw.fit(training_data, teachers)
results = scw.perdict(test_data)

teachers is 1-dimensional and training_data and test_data are 2-dimensional array.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

scw-1.0.tar.gz (2.4 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page