Classification_Data

source

Classification_Data()

Parameters

Parameter Name Type Range / Values Default Value Used for Description
Data Pandas Dataframe - - General Contains the dataset
test_size float 0.2-0.8 0.2 General Share/Percentage of Data used for testing, if pretrained model is used, all data (0.99) will be used for testing
x_labels list[str] headers from dataframe None General Labels used as evidence for the classification, if None all but y will be used
y_label str header from dataframe None General Label of column that contains the classes, if None final column will be used
hidden_layers array of ints [32]-[4096, 4096, 4096, 4096, 4096] [64, 64] Neural Net Nodes for each hidden layer, every entry in the array creates a hidden layer with as many nodes as the entry's value
training_epochs int 1 - 200 10 Neural Net
activation_func string elu, relu, linear, sigmoid, hard_sigmoid, softmax, softplus, tanh, exponential, gelu, selu, softsign, swish "relu" Neural Net https://www.tensorflow.org/api_docs/python/tf/keras/activations
validation_split Bool True Neural Net Whether during the training a part of the data will already be used for testing after each epoch, needed for accuracy/loss per epoch graphs
trees int 1 - 10.000 100 Random Forest Number of trees in the forest
model None General Allows user uploaded pre-trained models