Python load_data Examples

Example #1

File: fcn.py Project: akashlevy/Deep-Learn-Oil

import sys
import time
import qri
from keras.callbacks import EarlyStopping, ModelCheckpoint
from keras.layers.core import Activation, Dense, Dropout
from keras.models import Sequential
from keras.optimizers import SGD

# Model name
MDL_NAME = "base"

# Seed random number generator
np.random.seed(42)

# Load QRI data
datasets = qri.load_data("../datasets/qri.pkl.gz")

# Split into 2D datasets
train_set, valid_set, test_set = datasets

# Build neural network
model = Sequential()
model.add(Dense(48, 100, activation="relu"))
model.add(Dropout(0.5))
model.add(Dense(100, 12))

# Use stochastic gradient descent and compile model
sgd = SGD(lr=0.001, momentum=0.99, decay=1e-6, nesterov=True)
model.compile(loss=qri.mae_clip, optimizer=sgd)

# Use early stopping and saving as callbacks

Example #2

from keras.layers.core import Activation, Dense, Dropout
from keras.models import Sequential
from keras.optimizers import SGD
import sys

if len(sys.argv):
    first = int(sys.argv[1])

# Model name
MDL_NAME = "fcn-a-3hidden-thick%s" % first

# Seed random number generator
np.random.seed(42)

# Load QRI data
datasets = qri.load_data("qri.pkl.gz")

# Split into 2D datasets
train_set, valid_set, test_set = datasets

# Build neural network
model = Sequential()
model.add(Dense(48, first, activation="relu"))
model.add(Dropout(0.5))
model.add(Dense(first, first, activation="relu"))
model.add(Dropout(0.5))
model.add(Dense(first, first, activation="relu"))
model.add(Dropout(0.5))
model.add(Dense(first, 12))

# Use stochastic gradient descent and compile model

Example #3

File: cnn.py Project: vmolchan/Deep-Learn-Oil

import time
import qri
from keras.callbacks import EarlyStopping, ModelCheckpoint
from keras.layers.core import Activation, Dense, Dropout, Flatten
from keras.layers.convolutional import Convolution1D, MaxPooling1D
from keras.models import Sequential
from keras.optimizers import SGD

# Model name
MDL_NAME = "cnn"

# Seed random number generator
np.random.seed(42)

# Load QRI data
datasets = qri.load_data("../datasets/qri.pkl.gz")

# Split into 3D datasets
datasets = [(dataset[0][:, :, np.newaxis], dataset[1]) for dataset in datasets]
train_set, valid_set, test_set = datasets

# Build neural network
model = Sequential()
model.add(Convolution1D(1, 100, 13, activation="relu"))
model.add(Dropout(0.5))
model.add(Flatten())
model.add(Dense(3600, 12))

# Use stochastic gradient descent and compile model
sgd = SGD(lr=0.001, momentum=0.99, decay=1e-6, nesterov=True)
model.compile(loss=qri.mae_clip, optimizer=sgd)

Example #4

File: MultiLayerPerceptron.py Project: akashlevy/Deep-Learn-Oil

from keras.models import Sequential
from keras.optimizers import SGD
import sys

if len(sys.argv):
	first = int(sys.argv[1])
	

# Model name
MDL_NAME = "fcn-a-3hidden-thick%s"%first

# Seed random number generator
np.random.seed(42)

# Load QRI data
datasets = qri.load_data("qri.pkl.gz")

# Split into 2D datasets
train_set, valid_set, test_set = datasets

# Build neural network
model = Sequential()
model.add(Dense(48, first,activation="relu"))
model.add(Dropout(0.5))
model.add(Dense(first,first,activation="relu"))
model.add(Dropout(0.5))
model.add(Dense(first, first,activation="relu"))
model.add(Dropout(0.5))
model.add(Dense(first, 12))

# Use stochastic gradient descent and compile model