Skip to content

FDASTATAUBURN/fdnn-classification

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

62 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Deep Neural Network Classifier for Multi-dimensional Functional Data


Functional data pre-processing

  • Given functional data first equation, first use basis functions to extract projection scores second equation by integration.

Model input and output

  • Input: Projection scores xi.
  • Output: Binary class label k={-1, 1}.

Model selection

Neural network hyperparameters

  • J: number of projection scores for network inputs
  • L: number of layers
  • p: neurons per layer (uniform for all layers)
  • B: maximal norm of weights

Other hyperparameters

  • Loss function: hinge loss
  • Dropout rate: data dependent
  • Batch size: data dependent
  • Epoch number: data dependent
  • Activation function: ReLU
  • Optimizer: Adam

Function descriptions

One dimensional functional data

  • "dnn_1d_par.R": hyperparameter selection with training data. More details can be found in comments
  • "dnn_1d.R": functional deep neural netowrks. More details can be found in comments

Two dimensional functional data

  • "dnn_2d_par.R": hyperparameter selection with training data. More details can be found in comments
  • "dnn_2d.R": functional deep neural netowrks. More details can be found in comments

Examples

  • "example_1d.R": f, range, where psi1, psi2, psi3. Under class k, generate independently dis,
    where mu1, sigma1, mu2, sigma2.

  • "example_2d.R": f, range, where psi1, psi2, psi3, psi4. Under class k, generate independently dis,
    where mu1, sigma1, mu2, sigma2.

About

Functional classification via deep neural network

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages