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.
643
questions
1
vote
0answers
16 views
Ways of better regression performance - lower RMSE loss value [duplicate]
There's this dataset of the top 1000 streamers on Twitch at 2020. I'm currently solving a challenge problem, to predict the amount of Followers gained based on the other features of each channel.
The ...
0
votes
0answers
22 views
Is there a better way to solve this ML Problem?
I want to implement a neural network that estimates the scores in a game of a specific player. I already started and managed to build something that works kind of ok, but I see a lot of improvement ...
0
votes
0answers
17 views
Getting better results on LSTM (Tensorflow) shaping timesteps into features (?)
I am building multiple models for several different environments (similar applications) with LSTMs on Tensorflow. I know and I have seen people commenting on how inappropriate it is to build LSTMs ...
0
votes
0answers
6 views
Why loss function of my LSTM model drops a lot after training using the first 100 batches and then almost doesn't decrease?
I am training a LSTM model with one layer and 32 hidden neurons in tensorflow using tf.keras.layers.LSTMCell. The problem is a binary classification and the loss ...
1
vote
0answers
16 views
Dense layer with more neurons than on input
What is the purpose of having a dense layer with more neurons on the output than it received on input?
Let's imagine we have a neural network in which the last layer has the size of 1. Does it make ...
0
votes
0answers
25 views
Using GAN for generating data to augment training data
I want to model an experiment data using a neural network but as the data set size is too small (25 samples), I decided to use a GAN to generate more data. The input is a 4×1 tensor representing 4 ...
0
votes
0answers
12 views
What level prediction (node- or graph-level) is appropriate for my graph network problem?
I have experience in neural networks and just started exploring if graph NN is appropriate for my problem.
So, I have an undirected graph with nodes separated within a specified distance, say d, being ...
0
votes
0answers
11 views
How do I decide the steps per epoch?
I am training a segmentation model where training size is 4000 (768x768x3) images with a batch size of 4 images (because the GPU gives memory error above this).
My question or doubt is that when I am ...
0
votes
1answer
21 views
How does tf.keras.MultiHeadAttention layer handle positional encoding?
In Attention Is All You Need paper, positional encodings are added to the input embeddings in order to consider the order of the sequence. How does tf.keras.MultiHeadAttention handle positional ...
0
votes
0answers
5 views
Validation Loss gets better after adding Augmentation Layers but Test accuracy gets worse
I'm building a Siamese Network which should learn a face comparison function.
My model consists a CNN (which gets 2 inputs, and yields 2 embedding vectors). With the outputs I calculate:
...
1
vote
0answers
17 views
Cant understand training and validation accuracy [closed]
Im trying to train a model to classify some data in 6 categories. I get the following graphs using tensorboard
Im using Adam optimizer with learning rate 0.001 and I dont understand
Why the training ...
0
votes
0answers
10 views
How to interpret calibration curves in terms of class imbalance?
I am training a deep learning model toward a image classification task. The VGG-16 model is trained individually on two different training sets with varying degrees of data imbalance. Set-1 had only ...
0
votes
0answers
45 views
Precision, Recall and/or F1? Which should I use? or something different?
I am trying to use tensorflow to predict a decision based on a timeseries dataset.
I have three classes: "Wait", ...
0
votes
0answers
11 views
Feature importance of artificial neural networks with dependent features
I have trained a neural network model using TensorFlow and I would like to get feature importance of this model. I looked up some posts and found that ...
0
votes
0answers
5 views
Text Classification with BERT. Reddit posts. Accuracy is constant [duplicate]
I'm trying to clasify author age group ('under 21','young adult', 'adult). I'm working on this example: https://colab.research.google.com/github/tensorflow/text/blob/master/docs/tutorials/...
1
vote
0answers
22 views
Visualization of fully-connected neural networks
I'm trying to visualize a neural network built by keras with the following structure where all three Dense layers include a bias ...
0
votes
1answer
37 views
How Do You Know If A Problem Set Can Be Trained?
As the title suggests, how do you know if there exists a machine learning solution to a problem set? Earlier today I was working on building a neural network to predict whether stock prices will going ...
0
votes
0answers
12 views
Machine Learning Model Not Learning [duplicate]
I'm trying to build a LSTM model that takes in 150 consecutive candlesticks' open, high, low, close and EMA indicator values (150, 5), then predicts whether price moves 20 points up or down first.
<...
1
vote
1answer
45 views
How to deal with negative rewards in policy gradient with crossentropy loss
In policy gradient reinforcement learning we can use a loss function of the form -log(P)*reward, where P is the probability of ...
0
votes
0answers
16 views
How should i expect exponential decay to work in this case? (Adam optimizer)
I was facing high learning rate issues i.e., validation loss started to diverge after 9-13 epochs. In order mitigate that i have significantly reduced the learning rate from 4e-3 to 4e-4 and ...
0
votes
0answers
8 views
Why should I change the shape before sampling in a spatial variational Autoencoder?
When implementing a spatial variational autoencoder, I used this repo as a starting point. Before sampling, the author combines the two dimensions (see original code). Is there any reason to combine ...
0
votes
0answers
21 views
Is a convolution with $n$ input channels, $n$ filters and $n$ groups the same as depthwise convolution?
Let's say we have an input tensor $X$ with $n$ channels and a grouped convolution with $n$ groups and $n$ filters which produces an output tensor $Y$ with $n$ channels. Will this convolution ...
0
votes
0answers
27 views
Neural network initialization behavior: low initial loss that spikes, then slowly decreases
The embedded image is the training loss of my inception V3 architecture that I am training from scratch using the sigmoid cross-entropy loss function and the Tensorflow RMSProp optimizer with two ...
1
vote
0answers
27 views
Categorical Feature Embedding When a Specific Feature Can be Present in Multiple Fields
Goal: I own a restaurant that specializes in spaghetti, and I want to build a neural network that predicts overall customer satisfaction (output variable) of different spaghetti sauce recipes, based ...
1
vote
0answers
22 views
Multi-Multi-Class Classification
I'd like to build a model that can output results for several multi-class classification problems at once. Suppose you have diagnostic data about a product that needs to be repaired and you want to ...
0
votes
0answers
42 views
Why do all activation functions have positive slope?
I am wondering why all the common activation functions tend to increase with x (or stay flat like ReLU). I have not come across any that are inversely proportional to x, or that have some other shape. ...
0
votes
0answers
15 views
Reliability of tf.keras.metrics.AUC metric calculation
Please help understand how accurate or reliable tf.keras.metrics.AUC metric is.
It looks tf.keras.metrics.AUC is not actually ...
1
vote
1answer
23 views
Matrix to Matrix sequence architecture
My training data is the set of X{m x n} matricies, which have corresponding Y{m x n} output.It is like unknown function: Y = F(X) and I need to predict Y values based on X.
Cells of matrices can only ...
0
votes
0answers
26 views
Deep Reinforcement Learning, Target Networks Implementation
How does one implement target networks? I am using TensorFlow for deep reinforcement learning where a target network isn't currently implemented, and the training loop works as follows:
Get next ...
0
votes
0answers
19 views
Finding the best weights for sparse categorical cross entropy loss
In semantic segmentation and similar applications, sparse categorical cross entropy is often used as a loss function. Now it usually happens that samples are imbalanced. In my case, I have one class ...
2
votes
1answer
112 views
probability distribution as output for my LSTM
I have been trying to make a language model that predict the next word, but with the assumption that there are multiple "correct" answers.
Input: dictionary indices + document topic data for ...
0
votes
0answers
134 views
CIFAR-100 test accuracy maxes out at 67% but validation accuracy hits 90%
I've been experimenting with several different CNN models for CIFAR-100, but no matter what I do I cannot attain a test accuracy above 67%, even though validation accuracy sometimes reaches 90% for ...
0
votes
0answers
29 views
why reconstruction loss function multiplied by constant in VAE?
I try to understand best way how to use autoencoders loss functions.
So the often point is that common loss function consist of KL loss and reconstruction loss.
And what really confuse me is that ...
0
votes
0answers
43 views
Multi-output Neural Network only predicting one value
I have been using LSTM multi-output Neural Nets to perform two tasks, regression coupled with a classification. The data is in a time-series format where my dependent variable is trade quantity ...
0
votes
0answers
22 views
Best way to pre-process this data for a recommender?
I've created this data set, time series, with timestamps, and 3 columns, user_id, events, and next_hotel. I want to use next_hotel as a response variable.
user_id: ...
0
votes
0answers
21 views
Why my LSTM model predictions is a straigth line?
I am newbie in neural networks and I am trying to build a LSTM model to predict future values. My problem is that the plot of predictions result returns a line in comparation with the testting data.
I ...
1
vote
0answers
83 views
Can vanilla NN model a moving Average (MA) process?
Let say I simulate an AR(1) process. We can easily model the data with a vanilla NN with a single neuron as it is about finding the linear relationship between $y$ and $y_{t-1}$.
Now, how about a MA(1)...
3
votes
1answer
47 views
How are sentences one-hot encoded internally in an Embedding Keras Layer?
Multiple references are clear on how a single word is one-hot encoded in an Embedding layer, but what about sentences?
In order to illustrate an example, I will use the following SO reference. Let's ...
0
votes
0answers
15 views
Listwise Learning to rank for partially labelled list of ranking data
I was reading about the Listwise approach for LTR.
The first Listwise LTR paper ( refer page 3, left column, para 2-3) explicitly mentions "n(i)" which implies the number of ranked Documents ...
1
vote
1answer
65 views
Why does the residual 1x1 conv in wavenet not have an activation?
I have been trying to implement a wavenet. From the papers and designs I have looked at on github I have come up with the following...
...
2
votes
1answer
252 views
How to choose a batch size and the number of epochs while training a NN
After searching I read diferent theorys that using a greater batch size has better performance while model is training, but in the other hand, I also find the oposite view, that using a mini-batch ...
0
votes
0answers
53 views
Trained model returns constant value independent of input
I am trying to train a simple model on data that has seven inputs and one output. My dataframe basically looks like this:
...
0
votes
0answers
26 views
Is ther any logic that migth help while you create a Neural Network? [duplicate]
I am a newbie in Tensorflow and Keras and for the first moment I ask me the same few questions while I am creating my Neural Network. Even searching for some explanation of how to estructure, train, ...
0
votes
0answers
15 views
Why I get same predictions values for diferent input data?
I am newbie in the neural network world and actually I build my first neural network but for some reason when I use de trained model to predict ,giving by myselft the data I want it to predict it, ...
0
votes
0answers
13 views
How to check performance of the linear regression algorithm and improve?
Finally after countless tries I have successfully implemented Linear Regression using tensorflow on this dataset to predict Laptop prices after given specs. I plan to use this on a web app that can ...
0
votes
0answers
8 views
Passing multiple fixed length sequences to LSTM
The task at hand is multi-class text classification. I have 22,806 documents and 103 classes. The documents are of varying length:
...
0
votes
0answers
17 views
Why does my RandomizedSearchCV behave not only sub-optimally but also inconsistently?
This is a cross posting which has been left unanswered in Stack Overflow.
In the course of tuning the hyperparameters of a deep network for a (classic) classification problem, I am using ...
0
votes
0answers
24 views
Actionability from Tensorboard's weight histograms?
There are a few questions across the site about how to interpret the weight histograms. I understand what the histogram is showing, but:
What exactly does it mean when the weights in a given layer ...
0
votes
0answers
23 views
Music Generation Neural Network [duplicate]
So i'm working a machine learning project thats goal is to produce music. Now for the data, I am using librosa to load wav files. Basically the data produced by librosa is the sample rate of the song ...
0
votes
0answers
22 views
How to combine tabular data and images for neural network training? [duplicate]
I have a table in which there is an additional column with the image id:
...
