Regarding loss weighting in Keras regression problem for multiple outputs
0
I am running a Hyperas optimization for regression problem, with 3 predictors (X) and 2 targets (Y).
I did this, after ingesting the raw data:
X_train, X_val, Y_train, Y_val = train_test_split(X, Y, test_size=0.2, random_state=111)
# Input layers and Hidden Layers
model = Sequential()
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)]), input_dim = X_train.shape[1]))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)])))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
# Output layer
model.add(Dense(Y_train.shape[1]))
model.add(Activation('linear'))
model.compile(loss='mae', metrics=['mae'],optimizer=optimizer, loss_weights=[0.6,0.4])
history = model.fit(X_train, Y_train,
batch_size=choice([16,32,64,128]),
epochs=choice([20000]),
verbose=2,
validation_data=(X_val, Y_val),
callbacks=callbacks_list)
However, when running this, it says:
ValueError: When passing a list as loss_weights, it should have one entry per model output. The model has 1 outputs, but you passed loss_weights=[1, 1]
I'm guessing its due to the format of my inputs and outputs. However, I can't figure out the proper format for which I am supposed to feed it into the model.
Appreciate your advice please, thank you.
python keras
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
asked Nov 15 '18 at 13:22
CorseCorse
145110
add a comment |
0
I am running a Hyperas optimization for regression problem, with 3 predictors (X) and 2 targets (Y).
I did this, after ingesting the raw data:
X_train, X_val, Y_train, Y_val = train_test_split(X, Y, test_size=0.2, random_state=111)
# Input layers and Hidden Layers
model = Sequential()
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)]), input_dim = X_train.shape[1]))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)])))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
# Output layer
model.add(Dense(Y_train.shape[1]))
model.add(Activation('linear'))
model.compile(loss='mae', metrics=['mae'],optimizer=optimizer, loss_weights=[0.6,0.4])
history = model.fit(X_train, Y_train,
batch_size=choice([16,32,64,128]),
epochs=choice([20000]),
verbose=2,
validation_data=(X_val, Y_val),
callbacks=callbacks_list)
However, when running this, it says:
ValueError: When passing a list as loss_weights, it should have one entry per model output. The model has 1 outputs, but you passed loss_weights=[1, 1]
I'm guessing its due to the format of my inputs and outputs. However, I can't figure out the proper format for which I am supposed to feed it into the model.
Appreciate your advice please, thank you.
python keras
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
asked Nov 15 '18 at 13:22
CorseCorse
145110
Your model has one output layer and one loss function. So theloss_weightsdoes not make sense here, right? Don't confuse the the shape of output layer which is(2,)with the number of output layers. Loss functions are applied on the whole layers' output and not on each element of output layer individually.
– today
Nov 15 '18 at 13:27
ok so Sequential() cannot have more than one output layer right?
– Corse
Nov 15 '18 at 14:21
1
That's right. You need to use functional API instead if you want more flexibility.
– today
Nov 15 '18 at 14:24
alright, thank you!
– Corse
Nov 15 '18 at 14:36
add a comment |
0
0
0
I am running a Hyperas optimization for regression problem, with 3 predictors (X) and 2 targets (Y).
I did this, after ingesting the raw data:
X_train, X_val, Y_train, Y_val = train_test_split(X, Y, test_size=0.2, random_state=111)
# Input layers and Hidden Layers
model = Sequential()
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)]), input_dim = X_train.shape[1]))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)])))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
# Output layer
model.add(Dense(Y_train.shape[1]))
model.add(Activation('linear'))
model.compile(loss='mae', metrics=['mae'],optimizer=optimizer, loss_weights=[0.6,0.4])
history = model.fit(X_train, Y_train,
batch_size=choice([16,32,64,128]),
epochs=choice([20000]),
verbose=2,
validation_data=(X_val, Y_val),
callbacks=callbacks_list)
However, when running this, it says:
ValueError: When passing a list as loss_weights, it should have one entry per model output. The model has 1 outputs, but you passed loss_weights=[1, 1]
I'm guessing its due to the format of my inputs and outputs. However, I can't figure out the proper format for which I am supposed to feed it into the model.
Appreciate your advice please, thank you.
python keras
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
asked Nov 15 '18 at 13:22
CorseCorse
145110
I am running a Hyperas optimization for regression problem, with 3 predictors (X) and 2 targets (Y).
I did this, after ingesting the raw data:
X_train, X_val, Y_train, Y_val = train_test_split(X, Y, test_size=0.2, random_state=111)
# Input layers and Hidden Layers
model = Sequential()
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)]), input_dim = X_train.shape[1]))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
model.add(Dense(choice([np.power(2,1),np.power(2,2),np.power(2,3),np.power(2,4),np.power(2,5)])))
model.add(Activation(choice(['tanh','relu', 'sigmoid'])))
model.add(Dropout(uniform(0, 1)))
# Output layer
model.add(Dense(Y_train.shape[1]))
model.add(Activation('linear'))
model.compile(loss='mae', metrics=['mae'],optimizer=optimizer, loss_weights=[0.6,0.4])
history = model.fit(X_train, Y_train,
batch_size=choice([16,32,64,128]),
epochs=choice([20000]),
verbose=2,
validation_data=(X_val, Y_val),
callbacks=callbacks_list)
However, when running this, it says:
ValueError: When passing a list as loss_weights, it should have one entry per model output. The model has 1 outputs, but you passed loss_weights=[1, 1]
I'm guessing its due to the format of my inputs and outputs. However, I can't figure out the proper format for which I am supposed to feed it into the model.
Appreciate your advice please, thank you.
python keras
python keras
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
asked Nov 15 '18 at 13:22
CorseCorse
145110
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
asked Nov 15 '18 at 13:22
CorseCorse
145110
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
edited Nov 16 '18 at 9:45
Milo Lu
1,65711628
1,65711628
asked Nov 15 '18 at 13:22
CorseCorse
145110
asked Nov 15 '18 at 13:22
CorseCorse
145110
asked Nov 15 '18 at 13:22
CorseCorse
145110
145110
Your model has one output layer and one loss function. So theloss_weightsdoes not make sense here, right? Don't confuse the the shape of output layer which is(2,)with the number of output layers. Loss functions are applied on the whole layers' output and not on each element of output layer individually.
– today
Nov 15 '18 at 13:27
ok so Sequential() cannot have more than one output layer right?
– Corse
Nov 15 '18 at 14:21
1
That's right. You need to use functional API instead if you want more flexibility.
– today
Nov 15 '18 at 14:24
alright, thank you!
– Corse
Nov 15 '18 at 14:36
add a comment |
Your model has one output layer and one loss function. So theloss_weightsdoes not make sense here, right? Don't confuse the the shape of output layer which is(2,)with the number of output layers. Loss functions are applied on the whole layers' output and not on each element of output layer individually.
– today
Nov 15 '18 at 13:27
ok so Sequential() cannot have more than one output layer right?
– Corse
Nov 15 '18 at 14:21
1
That's right. You need to use functional API instead if you want more flexibility.
– today
Nov 15 '18 at 14:24
alright, thank you!
– Corse
Nov 15 '18 at 14:36
Your model has one output layer and one loss function. So the
loss_weights does not make sense here, right? Don't confuse the the shape of output layer which is (2,) with the number of output layers. Loss functions are applied on the whole layers' output and not on each element of output layer individually.– today
Nov 15 '18 at 13:27
Your model has one output layer and one loss function. So the
loss_weights does not make sense here, right? Don't confuse the the shape of output layer which is (2,) with the number of output layers. Loss functions are applied on the whole layers' output and not on each element of output layer individually.– today
Nov 15 '18 at 13:27
ok so Sequential() cannot have more than one output layer right?
– Corse
Nov 15 '18 at 14:21
ok so Sequential() cannot have more than one output layer right?
– Corse
Nov 15 '18 at 14:21
1
1
That's right. You need to use functional API instead if you want more flexibility.
– today
Nov 15 '18 at 14:24
That's right. You need to use functional API instead if you want more flexibility.
– today
Nov 15 '18 at 14:24
alright, thank you!
– Corse
Nov 15 '18 at 14:36
alright, thank you!
– Corse
Nov 15 '18 at 14:36
add a comment |
0
active
oldest
votes
Your Answer
StackExchange.ifUsing("editor", function ()
StackExchange.using("externalEditor", function ()
StackExchange.using("snippets", function ()
StackExchange.snippets.init();
);
);
, "code-snippets");
StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "1"
;
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function()
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled)
StackExchange.using("snippets", function()
createEditor();
);
else
createEditor();
);
function createEditor()
StackExchange.prepareEditor(
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader:
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
,
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
);
);
draft saved
draft discarded
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53320451%2fregarding-loss-weighting-in-keras-regression-problem-for-multiple-outputs%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
0
active
oldest
votes
0
active
oldest
votes
active
oldest
votes
active
oldest
votes
draft saved
draft discarded
Thanks for contributing an answer to Stack Overflow!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
draft saved
draft discarded
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53320451%2fregarding-loss-weighting-in-keras-regression-problem-for-multiple-outputs%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Your model has one output layer and one loss function. So the
loss_weightsdoes not make sense here, right? Don't confuse the the shape of output layer which is(2,)with the number of output layers. Loss functions are applied on the whole layers' output and not on each element of output layer individually.– today
Nov 15 '18 at 13:27
ok so Sequential() cannot have more than one output layer right?
– Corse
Nov 15 '18 at 14:21
1
That's right. You need to use functional API instead if you want more flexibility.
– today
Nov 15 '18 at 14:24
alright, thank you!
– Corse
Nov 15 '18 at 14:36