synthetic_aia_mia.predictor package¶
Submodules¶
synthetic_aia_mia.predictor.adult module¶
Define structures to manage and interface a fully connected neural network for adult.
- class synthetic_aia_mia.predictor.adult.AdultDataset(data)[source]¶
Bases:
Dataset
Pytorch dataset to handle adult data. :param data: dataset to convert. :type data: pandas.dataframe
- class synthetic_aia_mia.predictor.adult.AdultNN(overfit=False, loss='entropy', epochs=1000, hyper_sample=40, scale=True, tune=True)[source]¶
Bases:
object
Wrapper arround pytorch neural network. Interfare for hyper parameter optimisation using raytune.
- Parameters:
overfit (bool) – (Optional default=False) Force the model to overfit.
loss (string) – Used loss function. Either “entropy” or “mse”.
- fit(dadata)[source]¶
Train and tune hyper parameters.
- Parameters:
data – Dataset the will be split for training and hyper parameter tuning. Dataset must contain a column called “PINCP” used as training label.
- predict(dadata)[source]¶
Use a trained TabularNN to predict label of dataset.
- Parameters:
dadata (fetch_data.Dataset) – Dataset to evaluate.
- Returns:
Input dataset completed with hard labels, soft labels and loss.
- Return type:
- class synthetic_aia_mia.predictor.adult.TabularNN(input_size, l1, l2, l3, l4, output_size)[source]¶
Bases:
Module
Pytorch neural network for adult. :param input_size: Number of features. :type input_size: int :param hidden_size: Number of neurons/hidden layer. :type hidden_size: int :param l1: Size of the first layer. :type l1: int :param l2: Size of the second layer. :type l2: int :param l3: Size of the third layer. :type l3: int :param l4: Size of the fourth layer. :type l4: int :param output_size: Number classes in the labels. :type output_size: int
synthetic_aia_mia.predictor.utk module¶
Define structures to manage and interface a fully connected neural network for adult.
- class synthetic_aia_mia.predictor.utk.CNN(c1, c2, l)[source]¶
Bases:
Module
Convolutional neural network for 50x50x3 images.
- Parameters:
c1 (int) – Output number of channels of the first convolution layer.
c2 (int) – Output number of channels of the second convolution layer.
- Param:
Linear size.
- class synthetic_aia_mia.predictor.utk.UtkDataset(data)[source]¶
Bases:
Dataset
Pytorch dataset to handle StorageDataset. :param data: dataset to convert. :type data: fetch_data.utk.StorageDataset
- class synthetic_aia_mia.predictor.utk.UtkNN(epochs=500)[source]¶
Bases:
object
Wrapper arround pytorch neural network. Interfare for hyper parameter optimisation using raytune.
- Parameters:
epochs (int) – Number of epochs.
- fit(data)[source]¶
Train and tune hyper parameters.
- Parameters:
data – Dataset the will be split for training and hyper parameter tuning. Dataset must contain a column called “PINCP” used as training label.