Ggthjy

Why is mind meld hard for T'pol in Star Trek: Enterprise?

Who is this Ant Woman character in this image alongside the Wasp?

Is it a fallacy if someone claims they need an explanation for every word of your argument to the point where they don't understand common terms?

CREATE ASSEMBLY System.DirectoryServices.AccountManagement.dll without enabling TRUSTWORTHY

Why is working on the same position for more than 15 years not a red flag?

Do authors have to be politically correct in article-writing?

How to prevent users from executing commands through browser URL

How to count the characters of jar files by wc

awk + sum all numbers

What is the lore based reason that the Spectator has the Create Food and Water trait, instead of simply not requiring food and water?

Advice for a new journal editor

what does しにみえてる mean?

Porting Linux to another platform requirements

Intern applicant asking for compensation equivalent to that of permanent employee

Why publish a research paper when a blog post or a lecture slide can have more citation count than a journal paper?

Why do neural networks need so many training examples to perform?

How can my powered armor quickly replace its ceramic plates?

Can I string the D&D Starter Set campaign into another module, keeping the same characters?

Early credit roll before the end of the film

Can an insurance company drop you after receiving a bill and refusing to pay?

How would an AI self awareness kill switch work?

Pronunciation of umlaut vowels in the history of German

If I delete my router's history can my ISP still provide it to my parents?

Can I write a book of my D&D game?

Autoencoder for Dimensionality Reduction

Behavioral Differences between Standard Autoencoder and Variational AutoencoderRetain similarity distances when using an autoencoder for dimensionality reductionKeras LSTM: use weights from Keras model to replicate predictions using numpyautoEncoder as LSTM input, any benefit?How to set input for proper fit with lstm?Why is predicted rainfall by LSTM coming negative for some data points?What does SpatialDropout1D() do to output of Embedding() in Keras?Understanding LSTM structureAutoencoder Dimensionality ErrorLSTM sequence prediction: 3d input to 2d output

0

$begingroup$

My task is to reduce the features of my temporal sequence. Each input is of the shape (timesteps, features) = (240,117). I am using an autoencoder consisting of intermediate lstm layers and I am having difficulty for the same. I desire an output of shape (240,4).

The code for the same is

arr = x['Train_Data']



arr_list = []



for i in range(0,90):



    arr_list.append(arr[0,i])



arr = np.dstack(arr_list)



arr = np.rollaxis(arr,-1)



samples = arr.shape[0]



timesteps = arr.shape[1]



features = arr.shape[2]



model = Sequential()



model.add(LSTM(30, activation='relu', return_sequences = True, input_shape=(timesteps,features)))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(4, activation='relu', return_sequences=True))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(30, activation='relu', return_sequences = True))



model.add(LSTM(117, activation='relu', return_sequences=True))



model.add(TimeDistributed(Dense(features), input_shape=(timesteps,features)))



model.compile(optimizer='adam', loss='mse', metrics=['accuracy'])



model.fit(arr, arr, validation_split=0.33, epochs = 300)



model = Model(inputs=model.inputs, outputs=model.layers[2].output)



Train_Data_Reduced = model.predict(arr)

Am I going about it the right way? If I do not introduce a return_sequence = true for every layer then my output will be a fixed vector of the size(1,4) which is not what I desire. How do I go about it?

lstm autoencoder

share|improve this question

edited 7 hours ago

Hari_Sheldon

31

asked 13 hours ago

Sanjana KrishnamSanjana Krishnam

1

New contributor

Sanjana Krishnam is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.

$endgroup$

add a comment | 

0

$begingroup$

My task is to reduce the features of my temporal sequence. Each input is of the shape (timesteps, features) = (240,117). I am using an autoencoder consisting of intermediate lstm layers and I am having difficulty for the same. I desire an output of shape (240,4).

The code for the same is

arr = x['Train_Data']



arr_list = []



for i in range(0,90):



    arr_list.append(arr[0,i])



arr = np.dstack(arr_list)



arr = np.rollaxis(arr,-1)



samples = arr.shape[0]



timesteps = arr.shape[1]



features = arr.shape[2]



model = Sequential()



model.add(LSTM(30, activation='relu', return_sequences = True, input_shape=(timesteps,features)))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(4, activation='relu', return_sequences=True))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(30, activation='relu', return_sequences = True))



model.add(LSTM(117, activation='relu', return_sequences=True))



model.add(TimeDistributed(Dense(features), input_shape=(timesteps,features)))



model.compile(optimizer='adam', loss='mse', metrics=['accuracy'])



model.fit(arr, arr, validation_split=0.33, epochs = 300)



model = Model(inputs=model.inputs, outputs=model.layers[2].output)



Train_Data_Reduced = model.predict(arr)

Am I going about it the right way? If I do not introduce a return_sequence = true for every layer then my output will be a fixed vector of the size(1,4) which is not what I desire. How do I go about it?

lstm autoencoder

share|improve this question

edited 7 hours ago

Hari_Sheldon

31

asked 13 hours ago

Sanjana KrishnamSanjana Krishnam

1

New contributor

Sanjana Krishnam is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.

$endgroup$

add a comment | 

$begingroup$

My task is to reduce the features of my temporal sequence. Each input is of the shape (timesteps, features) = (240,117). I am using an autoencoder consisting of intermediate lstm layers and I am having difficulty for the same. I desire an output of shape (240,4).

The code for the same is

arr = x['Train_Data']



arr_list = []



for i in range(0,90):



    arr_list.append(arr[0,i])



arr = np.dstack(arr_list)



arr = np.rollaxis(arr,-1)



samples = arr.shape[0]



timesteps = arr.shape[1]



features = arr.shape[2]



model = Sequential()



model.add(LSTM(30, activation='relu', return_sequences = True, input_shape=(timesteps,features)))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(4, activation='relu', return_sequences=True))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(30, activation='relu', return_sequences = True))



model.add(LSTM(117, activation='relu', return_sequences=True))



model.add(TimeDistributed(Dense(features), input_shape=(timesteps,features)))



model.compile(optimizer='adam', loss='mse', metrics=['accuracy'])



model.fit(arr, arr, validation_split=0.33, epochs = 300)



model = Model(inputs=model.inputs, outputs=model.layers[2].output)



Train_Data_Reduced = model.predict(arr)

Am I going about it the right way? If I do not introduce a return_sequence = true for every layer then my output will be a fixed vector of the size(1,4) which is not what I desire. How do I go about it?

lstm autoencoder

share|improve this question

edited 7 hours ago

Hari_Sheldon

31

asked 13 hours ago

Sanjana KrishnamSanjana Krishnam

1

New contributor

Sanjana Krishnam is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.

$endgroup$

arr = x['Train_Data']



arr_list = []



for i in range(0,90):



    arr_list.append(arr[0,i])



arr = np.dstack(arr_list)



arr = np.rollaxis(arr,-1)



samples = arr.shape[0]



timesteps = arr.shape[1]



features = arr.shape[2]



model = Sequential()



model.add(LSTM(30, activation='relu', return_sequences = True, input_shape=(timesteps,features)))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(4, activation='relu', return_sequences=True))



model.add(LSTM(10, activation='relu', return_sequences = True))



model.add(LSTM(30, activation='relu', return_sequences = True))



model.add(LSTM(117, activation='relu', return_sequences=True))



model.add(TimeDistributed(Dense(features), input_shape=(timesteps,features)))



model.compile(optimizer='adam', loss='mse', metrics=['accuracy'])



model.fit(arr, arr, validation_split=0.33, epochs = 300)



model = Model(inputs=model.inputs, outputs=model.layers[2].output)



Train_Data_Reduced = model.predict(arr)

share|improve this question

edited 7 hours ago

Hari_Sheldon

31

asked 13 hours ago

Sanjana KrishnamSanjana Krishnam

1

New contributor

Sanjana Krishnam is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.

share|improve this question

edited 7 hours ago

Hari_Sheldon

31

asked 13 hours ago

Sanjana KrishnamSanjana Krishnam

1

New contributor

Sanjana Krishnam is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.

edited 7 hours ago

Hari_Sheldon

31

edited 7 hours ago

Hari_Sheldon

31

asked 13 hours ago

Sanjana KrishnamSanjana Krishnam

1

New contributor

Sanjana Krishnam is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.

asked 13 hours ago

Sanjana KrishnamSanjana Krishnam

1