Module rspamd_fann

This module enables fann interaction in rspamd Please note, that this module works merely if you have ENABLE_FANN=ON option definition when building rspamd

Brief content:

Functions:

rspamd_fann.is_enabled()

rspamd_fann.create(nlayers, [layer1, ... layern])

rspamd_fann.create_full(params)

rspamd_fann.load(file)

rspamd_fann.load_data(data)

rspamd_fann:data()

Methods:

rspamd_fann:get_inputs()

rspamd_fann:get_outputs()

rspamd_fann:get_mse()

rspamd_fann:get_layers()

rspamd_fann:save(fname)

Functions

The module rspamd_fann defines the following functions.

Function rspamd_fann.is_enabled()

Checks if fann is enabled for this rspamd build

Parameters:

No parameters

Returns:

• {boolean}: true if fann is enabled

Back to module description.

Function rspamd_fann.create(nlayers, [layer1, ... layern])

Creates new neural network with nlayers that contains layer1…layern neurons in each layer

Parameters:

• nlayers {number}: number of layers
• layerI {number}: number of neurons in each layer

Returns:

• {fann}: fann object

Back to module description.

Function rspamd_fann.create_full(params)

Creates new neural network with parameters:

• layers {table/numbers}: table of layers in form: {N1, N2, N3 … Nn} where N is number of neurons in a layer
• activation_hidden {string}: activation function type for hidden layers (tanh by default)
• activation_output {string}: activation function type for output layer (tanh by default)
• sparsed {float}: create sparsed ANN, where number is a coefficient for sparsing
• learn {string}: learning algorithm (quickprop, rprop or incremental)
• randomize {boolean}: randomize weights (true by default)

Parameters:

No parameters

Returns:

• {fann}: fann object

Back to module description.

Function rspamd_fann.load(file)

Loads neural network from the file

Parameters:

• file {string}: filename where fann is stored

Returns:

• {fann}: fann object

Back to module description.

Function rspamd_fann.load_data(data)

Loads neural network from the data

Parameters:

• file {string}: filename where fann is stored

Returns:

• {fann}: fann object

Back to module description.

Function rspamd_fann:data()

Returns serialized neural network

Parameters:

No parameters

Returns:

• {rspamd_text}: fann data

Back to module description.

Methods

The module rspamd_fann defines the following methods.

Method rspamd_fann:get_inputs()

Returns number of inputs for neural network

Parameters:

No parameters

Returns:

• {number}: number of inputs

Back to module description.

Method rspamd_fann:get_outputs()

Returns number of outputs for neural network

Parameters:

No parameters

Returns:

• {number}: number of outputs

Back to module description.

Method rspamd_fann:get_mse()

Returns mean square error for ANN

Parameters:

No parameters

Returns:

• {number}: MSE value

Back to module description.

Method rspamd_fann:get_layers()

Returns array of neurons count for each layer

Parameters:

No parameters

Returns:

• {table/number}: table with number ofr neurons in each layer

Back to module description.

Method rspamd_fann:save(fname)

Save fann to file named ‘fname’

Parameters:

• fname {string}: filename to save fann into

Returns:

• {boolean}: true if ann has been saved

Back to module description.

Back to top.