CRAN

OSTSC 0.0.1

Over Sampling for Time Series Classification

Released Dec 4, 2017 by Lan Wei

Dependencies

fields 9.6 foreach 1.4.4 MASS 7.3-50 doSNOW 1.0.16 doParallel 1.0.11

Oversampling of imbalanced univariate time series classification data using integrated ESPO and ADASYN methods. Enhanced Structure Preserving Oversampling (ESPO) is used to generate a large percentage of the synthetic minority samples from univariate labeled time series under the modeling assumption that the predictors are Gaussian. ESPO estimates the covariance structure of the minority-class samples and applies a spectral filer to reduce noise. Adaptive Synthetic (ADASYN) sampling approach is a nearest neighbor interpolation approach which is subsequently applied to the ESPO samples. This code is ported from a 'MATLAB' implementation by Cao et al. and adapted for use with Recurrent Neural Networks implemented in 'TensorFlow'.

Installation

Maven

This package can be included as a dependency from a Java or Scala project by including the following your project's pom.xml file. Read more about embedding Renjin in JVM-based projects.

<dependencies>
  <dependency>
    <groupId>org.renjin.cran</groupId>
    <artifactId>OSTSC</artifactId>
    <version>0.0.1-b7</version>
  </dependency>
</dependencies>
<repositories>
  <repository>
    <id>bedatadriven</id>
    <name>bedatadriven public repo</name>
    <url>https://nexus.bedatadriven.com/content/groups/public/</url>
  </repository>
</repositories>

View build log

Renjin CLI

If you're using Renjin from the command line, you load this library by invoking:

library('org.renjin.cran:OSTSC')

Test Results

This package was last tested against Renjin 0.9.2644 on Jun 2, 2018.

• OSTSC-examples
• input_data.OSTSC_handles_datasets_containing_NA,_NaN,_or_accidental_strings
• input_data.input_data_could_be_in_array_format
• input_data.input_data_could_be_in_data_frame_format
• input_data.input_data_could_be_in_xts_format
• testthat

Source

View GitHub Mirror

Release History

• 0.0.1