Questions tagged [neural-network]
Network structure inspired by simplified models of biological neurons (brain cells). Neural networks are trained to "learn" by supervised and unsupervised techniques, and can be used to solve optimization problems, approximation problems, classify patterns, and combinations thereof.
neural-network
20,224
questions
0
votes
0
answers
15
views
CrossEntropyLoss on PyTorch LSTM model with one classification per timestep
I am trying to make an LSTM model that will detect anomalies in timeseries data. It takes 5 inputs and produces 1 boolean output (True/False if anomaly is detected). The anomaly pattern will usually ...
- 1
0
votes
0
answers
9
views
Variable size input for CNN
I would like to create an AI that could guess the next image in a sequence of images and I have the idea to feed a CNN with 1,2,3,...,n-1 images as input running it through an LSTM and the output is ...
0
votes
0
answers
12
views
Big o notation of neural network
My problem is that how to calculate the computational complexity which used big o metrics of deep neural network,cart,lightbgm and random forest,
And where I can find the proof process of these?
I ...
- 1
0
votes
1
answer
32
views
How to solve CUDA error: device-side assert triggered?
for my thesis I'm working on Neural A* algorithm (https://github.com/omron-sinicx/neural-astar) and now my goal is use the algorithm on real maps provided by (https://github.com/micheleantonazzi/...
-1
votes
0
answers
17
views
Multi-class classification selecting machine learning model
In multiclass classification problem, will using neural network normally better than using svm model?
This is the plot I made with my data.
And will it be better for me to compute PCA and do the ...
- 1
-2
votes
0
answers
16
views
Not getting desired reconstructed image with python image denoising
I am making a program that makes a image noisy and then my program outputs a reconstructed image. I am using additive white gaussian noise and using a Feed Forward Neural Network. I have my program ...
- 29
-1
votes
0
answers
16
views
How to do feature and model selection with the parameter tuning in multi-class classification?
enter image description here
I want to use these three data, the sensor's distance from the ground ( since the distance will be different for a ground with ice and non-ice ), humidity, and surface ...
- 1
-3
votes
0
answers
16
views
Why we use concatenation in Unet? [closed]
why we use concatenation in unet and how did we get this grey lines. pls can someone elaborate this I'm new to this
I tried Unet architecture but I'm unable to understand how we get the grey lines i.e....
-1
votes
0
answers
12
views
Multi-label classification using bit-wise OR custom loss function
I have a question regarding this formula and the bit-wise OR and hope someone is familiar with the bit-wise OR for multi-label classification. Should I use for all R here the continuous output values, ...
- 109
-3
votes
0
answers
17
views
Binary classification: Accuracy always equal to 1 [closed]
training a CNN for a binary image classification task. However accuracy is always shown as 1 on validation and test scores. the loss curve does not show overfitting though.
tried to add class weights ...
- 1
0
votes
0
answers
29
views
Neural network codes in R [closed]
I am running my neural network codes in R for my survival data. I could run these codes but I need to calculate metrics (precision, accuracy, sensitivity, specificity) for the neural network model and ...
- 11
0
votes
0
answers
33
views
Input 0 of layer "dense" is incompatible with the layer: expected axis -1 of input shape to have value 36864, but received input with shape (190,1536)
I am trying to create a model for image classification for grayscale images. and got the error:
ValueError: Exception encountered when calling Sequential.call().
Input 0 of layer "dense" is ...
-1
votes
0
answers
28
views
Accuracy improving gets so slow after some epoches in neural network training [closed]
I have something like 70 million sample to train a neural network model, the accuracy improving goes very well and fast until somewhere like 25-30 epoch, after 25-30 epoch it gets so slow.
e.g.
epoch ...
- 1
-2
votes
0
answers
23
views
Choosing the Right Normalization Layer for Audio Classification Neural Network [closed]
I am building an audio classification neural network. I have learned that it is essential to normalize the data during network design.
However, I discovered that there are various types of ...
- 1
-3
votes
0
answers
31
views
Neural Network cost not changing [closed]
I can't seem to figure out why my cost print function is giving me the same cost every time and my weights and biases are not changing. Help would be greatly appreciated and I have tried putting it ...
