Package index • kindling

Model Specifications

Define neural network architectures for tidymodels

mlp_kindling(): Multi-Layer Perceptron (Feedforward Neural Network) via kindling
rnn_kindling(): Recurrent Neural Network via kindling
train_nnsnip() experimental: Parsnip Interface of train_nn()

Training Functions

Direct training interface (Level 2)

Generalized Neural Network Trainer

train_nn() experimental: Generalized Neural Network Trainer
nn_arch(): Architecture specification for train_nn()

Base Models

ffnn() rnn(): Base models for Neural Network Training in kindling

Utility Functions

nn_arch(): Architecture specification for train_nn()
early_stop(): Early Stopping Specification

Code Generators

Generate torch::nn_module() code (Lowest Level)

General-purpose / low-level generator, including the layer utilities & pronouns

nn_module_generator() experimental: Generalized Neural Network Module Expression Generator
.layer .i .in .out .is_output: Layer argument pronouns for formula-based specifications

High-level generators

ffnn_generator() rnn_generator(): Functions to generate nn_module (language) expression

Variable Importance

Interpret neural network models

garson(<ffnn_fit>) olden(<ffnn_fit>) vi_model(<ffnn_fit>): Variable Importance Methods for kindling Models

Tuning Parameters

Hyperparameter specifications for tidymodels

n_hlayers() hidden_neurons() activations() output_activation() optimizer() bias() validation_split() bidirectional(): Tunable hyperparameters for kindling models

Helper Functions

Utilities for model configuration

act_funs(): Activation Functions Specification Helper
args() superseded: Activation Function Arguments Helper
new_act_fn() experimental: Custom Activation Function Constructor
grid_depth(): Depth-Aware Grid Generation for Neural Networks
table_summary(): Summarize and Display a Two-Column Data Frame as a Formatted Table
ordinal_gen(): Ordinal Suffixes Generator