from data_load import load_CIFAR10
#from theano import function, config, shared, sandbox
#import theano.tensor as T
#
# from setup_GPU import setup_theano
#
# setup_theano()
from lasagne import layers
from lasagne.updates import nesterov_momentum
from lasagne.nonlinearities import softmax
from nolearn.lasagne import NeuralNet
cifar10_dir = "/home/kel/cifar-10-batches-py"
X_train, y_train, X_test, y_test = load_CIFAR10(cifar10_dir)
X_train_2d = np.dot(X_train[..., :3],
[0.299, 0.587, 0.114]).reshape(-1, 1, 32,
32).astype(np.float32)
X_test_2d = np.dot(X_test[..., :3],
[0.299, 0.587, 0.114]).reshape(-1, 1, 32,
32).astype(np.float32)
X_train_2d = (X_train_2d / 255.0) - 0.5
X_test_2d = (X_test_2d / 255.0) - 0.5
net2 = NeuralNet(
layers=[
('input', layers.InputLayer),
('conv1', layers.Conv2DLayer),
import tensorflow as tf
import numpy as np
import pandas as pd
import random
from data_load import load_CIFAR10
import matplotlib.pyplot as plt
from knn import knn
plt.rcParams['figure.figsize'] = (10, 8)
plt.rcParams['image.interpolation'] = 'nearest'
plt.rcParams['image.cmap'] = 'gray'
filepath = 'data/cifar-10-batches-py'
X_train, y_train, X_test, y_test = load_CIFAR10(filepath)
print('train data shape:', X_train.shape)
print('train labels shape:', y_train.shape)
print('test data shape:', X_test.shape)
print('test labels shape:', y_test.shape)
#########显示部分数据集
classes = [
'plane', 'car', 'bird', 'cat', 'deer', 'dog', 'drog', 'horse', 'ship',
'truck'
]
num_classes = len(classes)
samples_per_class = 7
for y, cls in enumerate(classes):
idxs = np.flatnonzero(y_train == y)
idxs = np.random.choice(idxs, samples_per_class, replace=False)
for i, idx in enumerate(idxs):
cbs_gauss_kmeans = [m.EarlyStopping(monitor='val_loss', patience=patience),
m.ModelCheckpoint(model_name_gauss_kmeans, monitor='val_loss',
verbose=0, save_best_only=True, mode='min')]
cbs_gauss_kmedoids = [m.EarlyStopping(monitor='val_loss', patience=patience),
m.ModelCheckpoint(model_name_gauss_kmedoids, monitor='val_loss',
verbose=0, save_best_only=True, mode='min')]
cbs_gauss_no_init = [m.EarlyStopping(monitor='val_loss', patience=patience),
m.ModelCheckpoint(model_name_gauss_no_init, monitor='val_loss',
verbose=0, save_best_only=True, mode='min')]
# Dataset Setup
if dataset == "CIFAR-10":
x_train, x_test, y_train, y_test = adl.load_CIFAR10()
n_classes = 10
else:
x_train, x_test, y_train, y_test = adl.load_CIFAR100()
n_classes = 100
x_train_pct, y_train_pct = m.sample_train(x_train, y_train, train_pct)
m.print_params(feature_extractor, embedding_dim, n_centers_per_class, n_classes,
lr, sigma, batch_size, epochs, dataset, input_shape, patience)
# Training Models
''' Softmax Model / Plain Model.
Without Initialization.
With Inverse Kernel.