Tagged Questions
0
votes
0answers
16 views
Implement an autoencoder flexible and stable with over 10 million iterations?
import numpy as np
import pandas as pd
import tensorflow as tf
n_inputs = 80
n_hidden = 32
n_outputs = n_inputs
learning_rate = 0.01
tf.reset_default_graph()
X = tf.placeholder(tf.float32, shape=...
1
vote
1answer
23 views
Does ELBO contain the reconstruction loss info in variational autoencoders
This is a related question to this - Variational autoencoder and reconstruction Log Probability vs Reconstruction error
I'm trying to understand how variational autoencoders are optimized. I've read ...
1
vote
1answer
32 views
Keras custom Loss function returning none
I am trying to code up an implementation of the variational autoencoder, however I am facing some difficulties regarding the loss function:
def vae_loss(sigma, mu):
def loss(y_true, y_pred):
...
0
votes
2answers
46 views
Keras Denoising Autoencoder (tabular data)
I have a project where I am doing a regression with Gradient Boosted Trees using tabular data. I want to see if using a denoising autoencoder on my data can find a better representation of my original ...
-1
votes
1answer
28 views
Questions about AutoEncoders
Good morning .
I have three questions about autoEncoders and i would really appreciate your help :
1- I have noticed that there is a lack of research papers done on deep auto encoders (AE) ...
-1
votes
0answers
31 views
Autoencoder that does NOT use MNIST
I'm working on a project for my undergrad Intro to AI class and have tasked with finding an autoencoder (I don't need to write my own, just modify somebody else's) that will take in our input. So far ...
2
votes
0answers
42 views
Keras custom regularization behavior using outputs of intermediate layers
I'm working on a Triplet Auto-encoder in Keras and I created a custom loss function that takes into account the reconstruction of the input and the values of the latent representation of the network.
...
-2
votes
0answers
11 views
Online Clustering with better representatives using Dimensionality reduction
I was trying to implement temporal online clustering for time-series data. General algorithms involve storing mediods/centroids for each cluster. I was thinking if we could get a better representative ...
0
votes
0answers
58 views
My autoencoder is bad in predicting but seems like a well trained one?
I developed a basic autoencoder of one layer to predict gene expressions where my input is a set of genes and their expression vector. I started by normalizing the dataset to -1 to 1 values for the ...
-1
votes
0answers
46 views
Trained pytorch auto-encoder model outputs junk on training subset inference
I'm trying to build a simple Auto-encoder with pytorch with the encoder being a standard convnet that outputs a latent code. The decoder is a reverse of the encoder that takes in the latent code and ...
0
votes
1answer
30 views
How to induce “uniform” sparsity/sparse coding in machine learning model?
I have a machine learning model (namely, an autoencoder) that attempts to learn a sparse representation of an input signal via a simple l1 penalty term added to the objective function. This indeed ...
0
votes
0answers
21 views
input shape and layer dimensions convolutional autoencoder
I want to train a convolutional Autoencoder with time series data to learn features and make later a prediction. I have the whole data in one dataframe and have 100 mio rows of data (1 column). From ...
0
votes
1answer
41 views
Calculating Gradient Update
Lets say I want to manually calculate the gradient update with respect to the Kullback-Liebler divergence loss, say on a VAE (see an actual example from pytorch sample documentation here):
KLD = -0.5 ...
0
votes
1answer
56 views
Deep autoencoder keeping constant accuracy in keras
I'm new in keras and deep learning field. In fact, I built a deep autoencoder using keras library based on ionosphere data set, which contains a mixed data frame (float, strings"objects", integers..) ...
-1
votes
1answer
71 views
Neural Networks - Difference between deep autoencoder and stacked autoencoder [closed]
Disclaimer: I also posted this question on CrossValidated but it is not receiving any attention. If this is not the place for it I will gladly remove it.
As I understand it, the only difference ...
7
votes
1answer
209 views
Intermediate layer makes tensorflow optimizer to stop working
This graph trains a simple signal identity encoder, and in fact shows that the weights are being evolved by the optimizer:
import tensorflow as tf
import numpy as np
initia = tf....
0
votes
0answers
135 views
Vector representation of time series in Keras stateful LSTM autoencoder
I am implementing a LSTM autoencoder in Keras to get a vector representation of my time series data.
The series I have are very long and so I am using stateful LSTMs.
I create non-overlapping ...
0
votes
1answer
141 views
Adapting the Keras variational autoencoder for denoising images
I'm trying to adapt the Keras example for VAE https://blog.keras.io/building-autoencoders-in-keras.html
I have modified the code to use noisy mnist images as the input of the autoencoder and the ...
0
votes
0answers
112 views
LSTM Encoder-Decoder as reconstruction model
I'm trying to build a LSTM Encoder-Decoder as a reconstruction model to detect anomalies in time-series data (similar to the one built in this paper here.) As mentioned in the paper, they used ...
2
votes
1answer
101 views
Variational Autoencoder cross-entropy loss (xent_loss) with 3D convolutional layers
I am adapting this implementation of VAE https://github.com/keras-team/keras/blob/master/examples/variational_autoencoder.py that i found here https://blog.keras.io/building-autoencoders-in-keras.html
...
1
vote
2answers
251 views
Keras autoencoder negative loss and val_loss with data in range [-1 1]
I am trying to adapt keras autoencoder example to a my data. I have the following network:
Xtrain = np.reshape(Xtrain, (len(Xtrain), 28, 28, 2))
Xtest = np.reshape(Xtest, (len(Xtest), 28, 28, 2))
...
-2
votes
1answer
50 views
How to extract feature names from reduced dimension set?
I am using PCA and/or Deep Autoencoders to reduce the dimensions of my set, but my question is instead of just having the reduced dimension of the set, is it possible or how is it possible to receive ...
2
votes
1answer
180 views
keras error in fit method : expected model_2 to have shape (None, 252, 252, 1) but got array with shape (300, 128, 128, 3)
I am building a image classifier for one-class classification in which i've used autoencoder.
While running this model I am getting this error by the line autoencoder_model.fit:
ValueError: Error ...
4
votes
2answers
295 views
ValueError: Error when checking target: expected model_2 to have shape (None, 252, 252, 1) but got array with shape (300, 128, 128, 3)
hi I am building a image classifier for one-class classification in which i've used autoencoder while running this model I am getting this error by this line (autoencoder_model.fit) (ValueError: Error ...
0
votes
1answer
111 views
keras autoencoder “Error when checking target”
i'm trying to adapt the 2d convolutional autoencoder example from the keras website: https://blog.keras.io/building-autoencoders-in-keras.html
to my own case where i use 1d inputs:
from keras.layers ...
0
votes
1answer
386 views
Keras - Autoencoder for Text Analysis
So I'm trying to create an autoencoder that will take text reviews and find a lower dimensional representation. I'm using keras and I want my loss function to compare the output of the AE to the ...
0
votes
1answer
98 views
From 2D to 3D using convolutional autoencoder
I’d like to reconstruct 3D object from 2D images.
For that, I try to use convolutional auto encoder. However, in which layer should I lift the dimensionality?
I wrote a code below, however, it shows ...
2
votes
0answers
645 views
How to prevent simple keras autoencoder from over compressing data?
I am trying to use the keras frontend with tensorflow backend for a simple autoencoder as a multidimensional scaling technique to plot multidimensional data into 2 dimensions. Many times when I run ...
0
votes
0answers
144 views
Anomaly score for anomaly detection
I would like to do the anomaly detection for multiple variables including categorical variable (activity) and continuous variable (cost) using an encoder-decoder architecture.
After training the ...
0
votes
0answers
122 views
Jointly training autoencoder for reconstruction and prediction accuracy in Keras
I'm trying to combine supervised and unsupervised learning using two different datasets. Suppose I want to solve a regression problem for which I have a small number K of labeled examples. At the same ...
0
votes
1answer
359 views
How to use masking layer to mask input/output in LSTM autoencoders?
I am trying to use LSTM autoencoder to do sequence-to-sequence learning with variable lengths of sequences as inputs, using following code:
inputs = Input(shape=(None, input_dim))
masked_input = ...
0
votes
1answer
61 views
how to properly get weights and biases from the model into the stacked autoencoder?
https://github.com/cmgreen210/TensorFlowDeepAutoencoder
i'm trying to save and restore model after the fine tuning step i tried to restore the model then get the variable from the model and it gave me ...
-1
votes
1answer
140 views
Why the loss of Variational Autoencoder in many implementations have opposite sign from paper?
I think I understand the paper of Auto-Encoding Variational Bayes. And I am reading some tensorflow codes implementing this paper. But I don't understand their loss function in those codes. Since lots ...
0
votes
1answer
206 views
TensorFlow tfrecords: tostring() changes dimension of image
I have built a model to train a convolutional autoencoder in TensorFlow. I followed the instructions on Reading Data from the TF documentation to read in my own images of size 233 x 233 x 3. Here is ...
0
votes
0answers
126 views
tensorflow autoencoder implementation produces lots of noise in reconstruction
As mentioned in the title, I implemented a stacked autoencoder using tensorflow to compress image data into lower dimensions. However, the reconstructed image, after learning finished, contains static ...
1
vote
1answer
490 views
Determining number of epochs for model fitting in Keras
I'm trying to automatically determine when a Keras autoencoder converges. For example, look at this link under "Let's build the simplest autoencoder possible." The number of epochs is hardcoded at 50 (...
0
votes
1answer
741 views
Classification with Keras Autoencoders
I'm trying to take a vanilla autoencoder using Keras (with a Tensorflow backend) and stop it when the loss value converges to a specific value. After the last epoch, I want to use a sigmoid function ...
-1
votes
2answers
211 views
data preprocessing tricks for auto-encoder
Recently, I try to use auto-encoder to find anomalies, but some of the input features are count data(e.g. number of clicks or number of shows). Do I need normalizing or scaling before training?
2
votes
1answer
135 views
Tensorflow major difference in loss between machines
I have written a Variational Auto-Encoder in Keras using Tensorflow as backend. As optimizer I use Adam, with a learning rate of 1e-4 and batch size 16. When I train the net on my Macbook's CPU (Intel ...
0
votes
0answers
15 views
Autoencoder sparsity enforces orthogonalization of the activations/encoding?
I'm not clear with this is question.
The sparsity penalization enforces orthogonality of the activations/encoding?
Best reggards,
0
votes
3answers
2k views
ValueError: Input 0 is incompatible with layer conv_1: expected ndim=3, found ndim=4
I am trying to make a variational auto encoder to learn to encode DNA sequences, but am getting an unexpected error.
My data is an array of one-hot arrays.
The issue I'm getting is a Value Error. ...
0
votes
0answers
276 views
TensorFlow Hidden Layer Content Autoencoder
I built a small Autoencoder to introduce myself into TF with 1 hidden Layer.
I created a database
X=[0010; 1010; 0110; 1110]
where the last 2 digits are unnecessary and should be dropped by the AE ...
4
votes
1answer
861 views
Keras Lambda layer has no output tensor shape, error when compiling model
I'm trying to build a character-level auto-encoder for text sequences using Keras. When I compile the model, I get an error concerning the shapes of the tensors as you can see bellow. I printed out ...
1
vote
2answers
2k views
How is a linear autoencoder equal to PCA?
I would like the mathematical proof of it. does anyone know a paper for it. or can workout the math?
1
vote
1answer
952 views
variable_scope causes 'variable does not exist' with Optimizer
Supoose I have two models foo and bar. Assume bar is pretrained and loaded. I want to define the cost function for foo as roughly sketched in the following code (it's actually an autoencoder). Please ...
1
vote
0answers
176 views
using validation monitor in tflearn.regression to create confusion matrix
So I have been trying to create a confusion metrics in my autoencoder
from __future__ import division, print_function, absolute_import
import numpy as np
#import matplotlib.pyplot as plt
import ...
1
vote
2answers
2k views
Keras autoencoder
I've worked a long time ago with neural networks in Java and now I'm trying to learn to use TFLearn and Keras in Python.
I'm trying to build an autoencoder, but as I'm experiencing problems the code ...
0
votes
1answer
308 views
Tensorflow autoencoder: How to get representative output?
Setup
I build an autoencoder with Tensorflow for images. My images are around 30 Pixels in length and width. I am using 5 layers:
Input layer
Encoder layer with 256 neurons with linear functions. (...
4
votes
0answers
1k views
Decoder's weights of Autoencoder with tied weights in Keras
I have implemented a tied weights Auto-encoder in Keras and have successfully trained it.
My goal is to use only the decoder part of the Auto-encoder as the last layer of another network, to fine ...
1
vote
0answers
228 views
Auto-encoder design tips
When training a stacked auto-encoder, what are the best practices to choose the number of hidden layers, and their sizes?
For example, is it best to set the hidden layer size to be one minus the ...
