Function reference • interpretnn

All functions

Boston: Standardized Boston Housing Data

covariate_eff(): Difference in average prediction for values above and below median

delta_method(): Perform delta method for a function FUN to calculate associated uncertainty

hessian(): Calculate hessian matrix

interpretnn() interpretnn.default() interpretnn.nnet() interpretnn.keras.engine.training.Model() interpretnn.nn() interpretnn.ANN() interpretnn.luz_module_fitted(): Statistically-Based Neural Networks

lr_test(): Likelihood ratio test for inputs

mlesim(): Perform m.l.e. simulation for a function FUN to calculate associated uncertainty

nnet_to_torch(): nnet weights to torch weights

nn_fit(): Fits various tracks (different random starting values) and chooses best model

nn_fit_nnet(): Fits various tracks (different random starting values) and chooses best model using nnet

nn_fit_torch(): Fits various tracks (different random starting values) and chooses best model using torch

nn_loglike(): Neural Network Normal Log-likelihood Value

nn_loss(): Neural network loss

nn_pred(): Neural network prediction

pce(): Partial covariate effect

plotci(): Plot Wald Confidence Intervals

plotnn(): Plot neural network architecture

sigmoid(): Sigmoid activation function

statnn-methods statnn,ANY-method statnn,keras.engine.training.Model-method: Methods for Function statnn in Package statnn

torch_to_nnet(): torch weights to nnet weights

VC(): Calculate variance-covariance matrix

wald_single_parameter(): Wald test for weights

wald_test(): Wald test for inputs