Dataset_Adiac: The automatic diatoms identification.

Description Usage Format Details Source

Description

The data is collected from a pilot study on automatic identification of diatoms (unicellular algae) from images.

Usage

1

Format

A dataset with 781 observations and a sequence length of 176, with a single sequence per row.

The y data is the class label (1 or 0).

The x data constructs time series sequences (numeric).

The training dataset contains 390 observations.

The testing dataset contains 391 observations.

Details

The dataset originally had 37 classes. This built-in data sets one class as the positive class (class 1) and all others are set to the negative class (class 0) to form a highly imbalanced dataset.

Source

http://timeseriesclassification.com/description.php?Dataset=Adiac


lweicdsor/OSTSC documentation built on May 8, 2019, 1:13 p.m.