Questions tagged [tensorflow]
A Python library for deep learning developed by Google. Use this tag for any on-topic question that (a) involves tensorflow either as a critical part of the question or expected answer, & (b) is not just about how to use tensorflow.
694
questions
1
vote
1
answer
22
views
Shape of tensorflow model input
I'm reading Masking and padding with Keras, in the beginning, an input example is:
...
- 2,121
0
votes
0
answers
8
views
TensorFlow, Lstm, integers [closed]
I have a data, that contains numerical and categorical features as well.
The labels i wish to predict are 4 different numbers, each one of them can be from 1 to 8(For example 1,4,7,4).
The label ...
0
votes
0
answers
10
views
how to determine epochs value, momentum and learning rate? [duplicate]
I have dataset with 70k-90k rows 42 columns and one of them is my target variable with 3 class label (yes, no , not applied). I'm applying sequential model (MLP) to my data, I just used random values ...
0
votes
0
answers
10
views
How to get correlated output from multiple tensorflow probability distributions
I'm building a tensorflow probability model that will output multiple (2+) predictions where the outputs will be draws from learned probability distributions. It currently looks something like this:
<...
- 61
0
votes
1
answer
26
views
Understanding keras layer structure/notation
I am trying to understand the following keras model:
...
- 101
1
vote
0
answers
24
views
Weighted Average of Multi Class AUC
Here, I can calculate the AUC score of each class individually in a multiclass problem (not to be confused with multilabel.)
...
0
votes
0
answers
19
views
Can the UNet architecture be adapted to perform image classification?
I am looking for ideas to adapt a UNet algorithm to perform image classification instead of segmentation. This is also in the context of transfer learning, as I already have pre-trained UNet models on ...
- 11
1
vote
0
answers
18
views
Data parallelism on multiple GPUs [closed]
I am trying to train a model using data parallelism on multiple GPUs. As I think, in data parallelism, we divide the data into batches, and then batches are deployed parallel. Afterward, the average ...
- 141
1
vote
1
answer
17
views
Use the Same Learning Rate to Train All Models When Doing Experiments For A Deep Learning Paper?
When writing a deep learning paper, I need to train several CNN models and compare their performances. They are from different architectures so different designs.
I'm wondering should I use the same ...
- 111
0
votes
0
answers
17
views
Getting normal distribution from spike distribution through data transformation
I'm trying to preprocess my data before feeding it to neural network. The goal is to get a normal distribution. I have 3 different features distributed as follows:
One can see bimodal, skewed and ...
1
vote
1
answer
22
views
Best Loss Function for Shape Resemblance in Time Series
Basically, predicting future values step by step using past values and some covariates as a feature, using some LSTM, Conv layers from tensorflow. I started by using mean absolute percentage error as ...
- 483
1
vote
0
answers
22
views
How we can interpret Playground on Tensorflow?
I need to understand how Playground on Tensorflow works.
How can I interpret x1, x2, x1^2, x2^2 and... . Because in real neural network, we just have some numbers ...
- 159
0
votes
0
answers
25
views
why the local reparameterization trick only works only for fully connected networks?
i was reading this article on towarddatascience and at a certain point the author says
"An important difference is that local reparametrization works only for fully connected networks, while ...
- 205
1
vote
1
answer
19
views
When adding batch norm layer do I need to added to all layers in DNN?
While developing deepfm model network I want to add batch norm layer because model seems to suffer from vanishing gradient. There are embedding layers, 2 layers a in deep model part and one dense ...
- 221
1
vote
0
answers
58
views
Why does the CNN model accuracy vary too much when the dataset is the same?
I have been working on a project where I have a lot of time series data(3000 csv file) from 6 different devices and I am trying to convert those data to an image array so that I can use them in CNN to ...
- 11
1
vote
0
answers
21
views
CycleGAN cycle loss
I was reading the paper of CycleGAN and I was trying to implement it.
However, my models does not converge to any good solution whatsoever, and since I've checked the implementation many times, I ...
1
vote
1
answer
30
views
How should I train my CNN with a tiny dataset
I'm working on a problem where I aim to classify sections of a track made on the floor using tape, into either left turns, right turns or straight track.
I'm struggling creating a CNN that is not ...
1
vote
0
answers
21
views
How to implement simple VAE with sparse tensor in Tensorflow
thank you for reading.
I have been attempting to train a simple VAE on very sparse 2D and 3D data. So far I have been training using dense tensors which - I think - is resulting in horrible training ...
- 11
0
votes
0
answers
20
views
Why is tensorflow with shap values only explain one feature on my dataset
I have a one dimensional convolutional network defined with tensorflow but now after I trained the model and done with all the evaluations I must explain feature contribution using SHAP Values. ...
- 1
0
votes
0
answers
72
views
Why is there a difference between training and validation accuracy when both of them are pointing to the same subset?
I was training a model and I accidentally pointed the training and testing set to the same dataset. I was surprised by the fact the validation and training accuracy are not the same. What could be the ...
2
votes
2
answers
51
views
Learning a simple pattern with RNN [duplicate]
I am trying to make RNN in tensorflow capture a basic pattern in a simple time series in hours. I am trying to solve a bigger problem involving count time series of customer demand.
The simple time ...
- 211
0
votes
0
answers
13
views
Preparing dataset over clients in federated learning
I am working on google cluster trace. I am working on small resource usage data only for 100 physical machine. Sample of data for two machine:
To predict the resource usage I used classical ANN. Now ...
1
vote
1
answer
72
views
Sequential Binary Imbalanced data classification with LSTM
I'm building an LSTM sequential Binary Classification Model, the data is highly imbalanced like say Fraud detection case.
After building an LSTM model on Sequential Vectorised data, I'm getting a very ...
0
votes
0
answers
49
views
Validation performance is really problematic; should I give up on increasing the validation performance of my deep learning model? [duplicate]
I am working on a multiclassification problem using time series data. Three datasets are utilized in this study. My deep neural network performs satisfactorily across two different data sets. However, ...
- 141
2
votes
1
answer
25
views
Tensorflow Playground - Explanation of the Feature colored with blue and orange
Tensorflow playground input feature $X_1$ is colored orange and blue. What is the meaning of these colors?
I think $(X_1, X_2)$ is training data point representing the coordinates of the dot to ...
- 866
0
votes
0
answers
8
views
Using optical flow to predict velocities
I am no expert in this field but more of a beginner with a bit of experience, so please keep the answer as simple as possible.
I cannot be very specific about this topic but what I am trying to do is ...
- 1
0
votes
0
answers
19
views
In what terms, training steps of my DG-GAN differs from the one of others?
Why does Deep Convolutional Neural Network GAN, trained using my step sequence, doesn't performs well as in the GFG DC-GAN Tutorial example? Even tho my code is similar in logic and resembles same to ...
0
votes
0
answers
273
views
Proper masking in MultiHeadAttention layer in Keras
I am new to Transformers and I am trying to create a very simple model (not NLP area) for processing data of variable length (not sequence data because for my problem order in data does not matter). ...
- 11
0
votes
0
answers
55
views
To understand the underlying math of GRU neural network in TensorFlow Keras
I tried out a simple GRU network with only 1 layer, 1 input tensor and 1 output, to verify its actual network connection (input nodes->hidden layer->output) by doing the manual calculation with ...
- 1
2
votes
2
answers
187
views
Is MSE loss a valid ELBO loss to measure?
I am learning from an example given by TensorFlow document, https://www.tensorflow.org/tutorials/generative/cvae#define_the_loss_function_and_the_optimizer:
VAEs train by maximizing the evidence ...
- 135
0
votes
1
answer
78
views
Training loss after last epoch differs from training loss (same data!) during evaluation
I am building a deep convolutional model with a custom loss function. As a first step, I am trying to bring training loss down as far as possible to see if my model can overfit.
Training with on only ...
0
votes
0
answers
15
views
Need Help for reward function in reinforcement learning
I've created a RL to trade on an artificial custom financial asset(Complete Code). This is my Dataframe(enviroment) made of 'Close' price and 'Volume':
...
- 1
0
votes
0
answers
80
views
Deformable Convolution implementation in tensorflow2
Is there any good implementation of Deformable Convolution in tensorflow2?
Deformable Convolutional Networks
Deformable ConvNets v2: More Deformable, Better Results
- 107
1
vote
0
answers
46
views
Detect exact position of a word or number in a sentence with machine learning
I'm trying to come up with an ML model/s to detect if a sentence has sensitive data(telephone number, IBAN, address, etc.) and also get the position of said sensitive data.
For Example "My name ...
- 11
0
votes
1
answer
63
views
How do implement a cross correlation as loss function? [closed]
I would like to use the normalized crosscorrelation coefficient NCC as a loss function in order to compare a output matrix A with a reference matrix B.
NCC=Sum_{ij} (A_{ij}-)(B_{ij}-)/(||A||*||B||)
...
MicSch
0
votes
0
answers
23
views
What kind of neural networks’ architecture is employed by Playground Tensorflow?
As you may know, Playground tensorflow allows you to design in real time an artificial NN and have a visual understanding of how it works.
How is it possible to have such a representation? Is that a ...
- 289
1
vote
0
answers
110
views
Classical VAE not learning 2D gaussian mixture distribution using MSE loss
I've been exploring VAE for non-image data. I consider small to medium-sized continuous vector spaces and I want to learn the distribution of a dataset in that space.
As a warm up exercise, I tried ...
- 113
0
votes
0
answers
24
views
Best way to approximate head point having only face keypoints
I'm using the BlazeFace model from TensorFlow which only has this few keypoints:
I need those keypoints plus a head keypoint, like this one:
My question is, which would be the best way to ...
0
votes
1
answer
154
views
How to properly mask MultiHeadAttention for sliding window time series data
I have data in the shape (batch, seq_len, features) that is a time series sliding window. In essence, I'm using the most recent ...
- 153
0
votes
0
answers
18
views
How to predict a mathematical progression with keras
I try the following model for a many-to-many recurrent network:
...
1
vote
0
answers
79
views
Tensorflow - calling a model inside a GradientTape scope VS calling it inside a loss function
Is there a difference in the gradient computation between the two code snippets ...
Code 1:
...
- 11
0
votes
0
answers
17
views
Updating temporal embeddings depending on the input
I'm building a forecasting model and I'm using a temporal embedding along with a positional embedding following the same architecture as Informer. ( https://arxiv.org/abs/2012.07436 )
My problem is ...
0
votes
0
answers
21
views
Predict gaps in time-series using LSTM
I have the number of timeseries with missing values (gaps). I want to train the LSTM NN for the prediction gaps task. Each time-series have the different numbers of gaps.
Now, I use the mean value ...
- 101
0
votes
0
answers
24
views
what is average unit activation and how to measure it?
recently I read a research paper about exponential linear units named "FAST AND ACCURATE DEEP NETWORK LEARNING BY EXPONENTIAL LINEAR UNITS (ELUS)". in section 4.1.1 they train 8 layers (128 ...
1
vote
1
answer
102
views
Train a Final Machine Learning Model with Tensorflow
Based on a previous question and on this article, it is suggested that you split the data between train and test (or train/validate/test). But once you have control of your model, you should retrain ...
- 147
0
votes
0
answers
36
views
Understanding tf.feature_column and dimensionality input
I hope this is the correct forum.
I am going over the feature_column package of the TensorFlow [1] and have checked the code that generates a DNN using the feature_column. Assume that there is a ...
- 111
0
votes
0
answers
39
views
Neural Network Regression Making Bad predictions
I have a dataset that has 83 dimentions and 300K observations for the training set. These 300K observations are standardized and fitted to a 3-layer DNN with the following parameters:
...
- 3
0
votes
0
answers
9
views
Image-recognition model makes good predictions only with training examples
Im trying to use a kaggle dataset to train a model that recognizes american fingerspelling language from an image. The problem is that, built the model, if i record the screen with the examples ...
- 1
5
votes
1
answer
365
views
Neural network gives very different accuracies if repeated on same data, why?
I'm running a neural network to classify audio files into 4 classes. This uses 3300 1min files split roughly evenly across classes. I split this into 80:10:10 train:validation:test sets. This trains ...
- 63
0
votes
0
answers
594
views
Choosing the 'best' epoch to stop the training of neural network. Top accuracy not improving, but average is
I'm familiar with concepts like early stopping, and detection of plateau and so on.
Tensorflow CNN training has a possibility of saving only best model too, according to model's accuracy metric (for ...
- 101