def basic_model():
"""
TODO:
Initialize a classifier
Suggested classifiers:
* SVM
* K-nn
* Decision Tree
* Random Forrest
Train a model and evaluate it
You should use k-fold cross validation
Return the accuracy of each fold in the scores variable
"""
model = None
"""
labels: a list with a subject-label for each feature-vector
targets: the class labels
features: the features
"""
labels, targets, features = data.get_all("features.csv")
features = data.normalize(features)
scores = None
return scores
def basic_model():
model = RandomForestClassifier(n_estimators=100)
labels, targets, features = data.get_all("features.csv")
features = data.normalize(features)
scores = cross_val_score(model, features, targets, cv=5)
return scores
def personalized_model():
model = RandomForestClassifier(n_estimators=100)
labels, targets, features = data.get_all("features.csv")
logo = LeaveOneGroupOut()
scores = []
for rest, user in logo.split(features, targets, groups=labels):
normalized = data.normalize(features[user])
user_scores = cross_val_score(model, normalized, targets[user], cv=5)
scores = scores + user_scores.tolist()
return np.array(scores)
def personalized_model():
"""
TODO:
Initialize a classifier
Train a model and evaluate it
Here you will train one model per subject in the labels-list
Remember to normalize the features for each subject seperately.
Evaluate each model using k-fold cross validation
Return the accuracy of each fold for every user in the scores variable as a numpy-array
"""
model = None
labels, targets, features = data.get_all("features.csv")
scores = None
return np.array(scores)
def general_model():
"""
TODO:
Initialize a classifier
Train a model and evaluate it
Here you will use leave-one-group-out cross validation to create a general model
Remember to normalize the training data set and the test data set seperately.
There is a function to do this in data.py
Return the accuracy of each fold in the scores variable as a numpy-array
"""
model = None
labels, targets, features = data.get_all("features.csv")
scores = None
return np.array(scores)
def general_model():
model = RandomForestClassifier(n_estimators=100)
labels, targets, features = data.get_all("features.csv")
logo = LeaveOneGroupOut()
scores = []
for train_index, test_index in logo.split(features, targets,
groups=labels):
features_copy = np.copy(features)
X_train, X_test = data.normalize_train_test(features_copy[train_index],
features_copy[test_index])
y_train, y_test = targets[train_index], targets[test_index]
model.fit(X_train, y_train)
user_score = model.score(X_test, y_test)
scores.append(user_score)
return np.array(scores)
def ask_basic_info(library):
try:
info = data.get_all()
library_id = request.intent.slots.library.resolutions.resolutionsPerAuthority[
0]['values'][0]['value']['id']
for library_info in info:
if library_info['unit_id'] == int(library_id):
email = library_info['contact_email']
answer_msg = render_template("answer-basic-info",
library=library,
info=library_info)
break
# No library matching / date matching
except (KeyError, AttributeError) as e:
answer_msg = render_template("error-no-match")
# other errors
except:
answer_msg = render_template("error-other")
return question(answer_msg)
# Cuda
use_gpu = torch.cuda.is_available()
if use_gpu:
print("Using gpu")
else:
print('Using cpu')
# embedding path
emb_path = '/p300/TensorFSARNN/data/emb/glove.6B/glove.6B.100d.txt'
# load data
dir = '/p300/MerryUp_RE2RNN/dataset/ATIS/'
train_tokens, train_tags, train_seqlen = data.load_data_ATIS(dir + 'train.json')
test_tokens, test_tags, test_seqlen = data.load_data_ATIS(dir + 'test.json')
# get all
all_tokens, all_tags = data.get_all(train_tokens, test_tokens, train_tags, test_tags)
# get max
max = data.get_maxlen(all_tokens)
# get dict
word2idx, idx2word, tag2idx, idx2tag = data.create_dict(all_tokens, all_tags)
vocab_size = len(word2idx) # vocab_size = 942
tag_size = len(tag2idx) # C = 27
# add padding
train_tokens = data.add_padding(train_tokens, max)
test_tokens = data.add_padding(test_tokens, max)
# convert2vec
train_tokens, train_tags = data.convert2vec(train_tokens, train_tags, word2idx, tag2idx)
test_tokens, test_tags = data.convert2vec(test_tokens, test_tags, word2idx=word2idx, tag2idx = tag2idx)
# dataset
train_dataset = myDataSet(train_tokens, train_tags, train_seqlen)
test_dataset = myDataSet(test_tokens, test_tags, test_seqlen)