Single Encoder Autoencoder

The single-encoder autoencoder has a single encoder and multiple layers of decoder. This makes it suitable for simple data representations.

from sdgne.datagenerator.autoencoder import SingleEncoderAutoencoder

minority_column_label = 'class'
minority_class_label = 0

synthesizer = SingleEncoderAutoencoder(dataset,
                               minority_column_label,
                               minority_class_label)

Importing Heavy Decoder Autoencoder

from sdgne.datagenerator.autoencoder import SingleEncoderAutoencoder

---

Creating a synthesizer

synthesizer = SingleEncoderAutoencoder(dataset,
                               minority_column_label,
                               minority_class_label)

Parameters

dataset	required	pd.Dataframe	Represents a pandas data frame containing both, the original minority and the original majority data
minority_column_label	required	string	Represents the column label. Eg. 'class' , 'output'
minority_class_label	required	string	Represents the minority class label. Eg. '1', '0'

Returns

An instance of class SingleEncoderAutoencoder.

Network Architecture

Below is the network architecture for the Single Encoder Autoencoder. Here, the encoder has a single layer, with 20 nodes. The bottleneck is a single Dense layer with 16 nodes. The decoder has two layers with 18 and 20 nodes respectively. The decoder layer activation is sigmoid.

encoder_dense_layers	20
bottle_neck	16
decoder_dense_layers	18, 20
decoder_activation	sigmoid

Synthetic Data Generation

data_generator()

The data generator function generates synthetic data using the synthesizer. It has an option parameter no_of_syntetic_data.

• If no_of_syntetic_data is not defined, data_generator by default generates the `n` number of synthetic data such that the majority data and minority datasets get balanced.

• If no_of_syntetic_data is defined and the dataset is already balanced, data_generator generates 2 * Number of original minority data.

• If no_of_syntetic_data is defined and the dataset is not balanced, it generates synthetic data equal to the value passed.

synthesized_data = synthesizer.data_generator(no_of_syntetic_data)

Parameters

no_of_syntetic_data	optional (default: None)	integer	Represents the number of synthetic data to be generated

Returns

A pandas data frame that combines original data and synthetic data.

Usage

synthesizer.data_generator()

synthesizer.data_generator(no_of_syntetic_data=100)

synthesize_data returned from data_generator( ) adds a column `synthetic_data` to the data frame.

df['synthetic_data'] = 0 : For original data df['synthetic_data'] = 1 : For Synthetic generated data