manifest.json

{
  "title": "ImageEmbeddingCAE",
  "description": "Implementation of http://arxiv.org/abs/2009.02040",
  "inputDimensionality": "univariate",
  "version": "0.3.0",
  "authors": "Gabriel Rodriguez Garcia, Gabriel Michau, Melanie Ducoffe, Jayant Sen Gupta",
  "language": "Python",
  "type": "Detector",
  "learningType": "Semi-Supervised",
  "mainFile": "algorithm.py",
  "trainingStep": {
    "parameters": [
      {
        "name": "anomaly_window_size",
        "type": "int",
        "defaultValue": 512,
        "optional": "true",
        "description": "length of one time series chunk (tumbling window)"
      },
      {
        "name": "kernel_size",
        "type": "int",
        "defaultValue": 2,
        "optional": "true",
        "description": "width, height of each convolution kernel (stride is equal to this value)"
      },
      {
        "name": "num_kernels",
        "type": "int",
        "defaultValue": 64,
        "optional": "true",
        "description": "number of convolution kernels used in each layer"
      },
      {
        "name": "latent_size",
        "type": "int",
        "defaultValue": 100,
        "optional": "true",
        "description": "number of neurons used in the embedding layer"
      },
      {
        "name": "leaky_relu_alpha",
        "type": "float",
        "defaultValue": 0.03,
        "optional": "true",
        "description": "alpha value used for leaky relu activation function"
      },
      {
        "name": "batch_size",
        "type": "int",
        "defaultValue": 32,
        "optional": "true",
        "description": "number of simultaneously trained data instances"
      },
      {
        "name": "test_batch_size",
        "type": "int",
        "defaultValue": 128,
        "optional": "true",
        "description": "number of simultaneously trained data instances"
      },
      {
        "name": "learning_rate",
        "type": "float",
        "defaultValue": 0.001,
        "optional": "true",
        "description": "Gradient factor for backpropagation"
      },
      {
        "name": "epochs",
        "type": "int",
        "defaultValue": 30,
        "optional": "true",
        "description": "number of training iterations over entire dataset"
      },
      {
        "name": "split",
        "type": "float",
        "defaultValue": 0.8,
        "optional": "true",
        "description": "train-validation split"
      },
      {
        "name": "early_stopping_delta",
        "type": "float",
        "defaultValue": 0.05,
        "optional": "true",
        "description": "If 1 - (loss / last_loss) is less than `delta` for `patience` epochs, stop"
      },
      {
        "name": "early_stopping_patience",
        "type": "int",
        "defaultValue": 10,
        "optional": "true",
        "description": "If 1 - (loss / last_loss) is less than `delta` for `patience` epochs, stop"
      },
      {
        "name": "random_state",
        "type": "int",
        "defaultValue": 42,
        "optional": "true",
        "description": "Seed for the random number generator"
      }
    ],
    "modelInput": "none"
  },
  "executionStep": {
    "parameters": [
      {
        "name": "random_state",
        "type": "int",
        "defaultValue": 42,
        "optional": "true",
        "description": "Seed for the random number generator"
      }
    ],
    "modelInput": "required"
  }
}