Questions tagged [neural-networks]

Artificial neural networks (ANNs) are a broad class of computational models loosely based on biological neural networks. They encompass feedforward NNs (including "deep" NNs), convolutional NNs, recurrent NNs, etc.

Filter by
Sorted by
Tagged with
0
votes
0answers
3 views

Two True Positive for one ground truth in object detection

I am wondering is it possible to have two true positive predictions for one bounding box ground truth only. Following this section from Stanford. They define truth positive like this: We start with ...
4
votes
3answers
646 views

Forecasting Prices vs Returns by Deep Learning

The question: It seems that (univariate) forecasting stock market done by websites using DL and LSTM actually does not work that well if we focus on returns instead of prices. What is a relatively ...
0
votes
0answers
7 views

NN and Output in two case, how this solution is reached?

I read my notes and get into stuck here. is there any idea how $5$ and $3$ is calculated here for option $1$ and option $2$?
0
votes
0answers
24 views

Using keras in R to perform neural network, my model has very low accuracy but the prediction is good and I don't know why

I used the classic dataset - fashion mnist dataset that has 784 columns of pixels and 1 column of the label (from 0 to 9), and I was going to transform the images into their corresponding seven ...
1
vote
1answer
525 views

Can attention be implemented without encoder / decoder?

I just got into models beyond biLSTM, would like to start with applying attention to my existing network (RNN). I find examples for attention always with encoder decoder architecture, however is it ...
0
votes
0answers
8 views

Post-hoc analysis of neural network predictions

I have trained a model in PyTorch. My model predicts results of football games. I hypothesize that certain games in my test-set will have higher accuracy. One of the variables would be the start of ...
0
votes
0answers
8 views

Is there a “good” way to evaluate segmentation in weakly labeled image data?

I have an image dataset for anomaly detection, which has weakly labeled ground truth images for the anomalies. Therefore, if there is a defect in an image, the ground truth would have a relatively big ...
2
votes
1answer
194 views

How to feed multivariate spatio-temporal data into cnn?

After trying to find an example for quite a while, I finally came to ask my question here: What I have: I have a temporal sequence of 2d spatial data with 100 cells(or pixels) in longitude and 30 ...
0
votes
0answers
7 views

N-to-1 frame CNN prediction model: how to represent the data?

I’m relatively new to vision modeling and I have a PyTorch CNN model that can predict the next frame of image based on 1 input frame relatively well. Now I want this model to output 1 image based on a ...
0
votes
2answers
25 views

Different input size for training and prediction in CNN for image segmentation?

I’m relatively unexperienced when it comes to deep learning and I’m trying to reimplement a CNN architecture for segmentation of medical images based on a paper. In the paper they state that they use ...
0
votes
0answers
14 views

How to calculate 28 day mortality?

I have a retrospective EHR database from a hospital and I would like to build an ML model to predict whether a patient will die within 28 days or not (from discharge/some time point T) Can I check ...
2
votes
1answer
99 views

Which activation function is better to a 1-dimensional time series in a LSTM model?

I am experimenting with a LSTM model (I have normalized the data) until now the 'sigmoid' performs better than others. How can I justify/interpret it?
0
votes
1answer
218 views

Does weights learned from pre-training using a Denoising Autoencoder needs rescaling when using dropout in complete NN

I have a question related to this question which is also yet to be answered. I am using a denoising autoencoder for pre-training of a neural network for dimensionality reduction. I want to use the ...
87
votes
5answers
75k views

Where should I place dropout layers in a neural network?

Is there any general guidelines on where to place dropout layers in a neural network?
2
votes
1answer
49 views

Why the Gaussian Distribution in ML/DL?

So whenever I learn about a method, or a technique to combat some disadvantage present in said method, some mention of transforming data into the Gaussian Distribution exists. My focus here is mainly ...
0
votes
0answers
12 views

Machine learning methods for filtering noisy data

I have my data (hear beat) cleaned up with the appropriate filter, however, they are still noisy and need to extract a clean pattern. Which is the best algorithm for this purpose? I have also clean ...
0
votes
1answer
16 views

Optimize number of hidden layers and neurons with RandomizedSearchCV (scikit-learn) -> No unnecessary trainings?

I want to optimize the number of hidden layers and the number of units in each hidden layer. For this I used RandomizedSearchCV from sklearn in this way: ...
1
vote
1answer
37 views

Why the length of the weight vector is likely to grow?

In the book Michael Nielsen's Neural Networks and Deep Learning, in chapter 3 he writes: "Heuristically, if the cost function is unregularized, then the length of the weight vector is likely to ...
1
vote
1answer
425 views

Q-Learning using ANN with continous action and variable-length state

The question is basically What should I do if state vector has variable length? If the action is bounded and continuous, how can I obtain max(Q(state,action)) without using painfully slow global ...
2
votes
3answers
5k views

Why do autoencoders come under supervised learning?

Autoencoder is an unsupervised learning method. How? I have searched / read many documents, they mention it (autoencoder) as unsupervised learning, but there is no answer how it is?
4
votes
1answer
250 views

Does Attention Help with standard auto-encoders

I understand the use of attention mechanisms in the encoder-decoder for sequence-to-sequence problem such as a language translator. I am just trying to figure out whether it is possible to use ...
1
vote
0answers
8 views

why 1*1 layer after max-pooling in inception network

I understand that in inception network, 1 * 1 layer is used before 3 * 3 or 5 * 5 filter to do some channel reduction and make computation easier. But why max-pooling then 1 * 1 layer? In particular, ...
1
vote
0answers
13 views

Is it a good idea to continue training a model after the train/validation accuracy has stopped improving?

The following animated diagram shows the training statistics of a Deep Neural Network classifier at the end of each epoch: The diagrams on the left show the accuracy (upper) and loss (lower) values ...
3
votes
2answers
346 views

How to model an “order-invariant” function by neural networks

I want to approximate a multi-variable function $f(x_1,x_2,x_3,x_4,x_5,y)$ from data by neural networks, and $f$ satisfies $f(x_1,\ldots,x_5,y)=f(x_{i_1},\ldots,x_{i_5},y)$, where $(i_1,\ldots,i_5)$ ...
0
votes
1answer
23 views

How to design a VAE for a given image dataset?

I have a dataset with texture images of the size 512x512. I tried to train a VAE with 3 "blocks" (Convolution , Activation and Batchnormalization) in the encoder and in the decoder and a ...
1
vote
1answer
371 views

Regression to find optimal linear combination of ensemble of neural network weightings?

I have N neural networks trained on different subsets of features of a dataset and with slightly different methods. My problem is a multiclass output, with the output layer comprising of the softmax ...
1
vote
0answers
22 views

LSTM: Best way to handle categories, in practice

There is an issue emerging in the practical use of Long-Short-Term-Memory (LSTM) Deep Neural Nets (DNN) with my use case. In typical machine learning scenarios one encounters in benchmark datasets, or ...
0
votes
0answers
20 views

Can a neural network still manage to converge, with slightly incorrect gradients?

In a network, we find gradients of the error function w.r.t each of the parameters used in the network. We then update the weights say, using vanilla Gradient Descent. If the computed gradients, do ...
0
votes
0answers
6 views

How to interpret 0 in the context of a neuron's value in a neural network

I've noticed there are two ways to interpret 0 and I'm a bit confused. Interpretation #1 A 0 is just as meaningful as any non-zero number. Examples and reasoning: When we normalize images for input ...
0
votes
0answers
8 views

How to predict an event for different time intervals and compute score?

Let's say I have a medical dataset/EHR dataset that is retrospective and longitudinal in nature. Meaning one person has multiple measurements across multiple time points (in the past). This dataset ...
7
votes
0answers
71 views

Overfitting and underfitting in Neural Networks: Is total number of neurons or number of neurons per layer more relevant?

I have seen posts where the discussion was centered around the effect of big and small total number of neurons in a neural network, especially with respect to the potential of the network to overfit ...
2
votes
4answers
349 views

Loss function for spam detection like problems

I am working on a deep learning problem where wrong classifications of fake events are not problematic, but where the opposite case is very problematic. I suppose this is similar to how spam detectors ...
1
vote
0answers
18 views

Order of dropout and activation in 1D convolutional networks

I have a simple cnn-lstm network. There are two 1D convolutional layers after the input layer. Every 1D convolutional layer is followed by a dropout. What I observe is that when I have conv1D -> ...
2
votes
0answers
46 views
+100

Metrics for implicit data in the recommender system with NCF

Which metrics for analysis and evaluation for implicit data in a recommender system do you use? And which ones do you use when you are looking for the closest neighbors to make a recommendation? I'm ...
0
votes
0answers
6 views

Optimally stratifying training, validation and testing samples using simulated targets

I'm fitting a large-scale (both in size of sample and input vector) single hidden-layer feedforward neural network on simulated targets, $\tilde{t}_{\tilde{n}} \in \{\tilde{t}_1,\cdots,\tilde{t}_\...
56
votes
4answers
58k views

Why is logistic regression a linear classifier?

Since we are using the logistic function to transform a linear combination of the input into a non-linear output, how can logistic regression be considered a linear classifier? Linear regression is ...
1
vote
0answers
16 views

Why conditional GAN does not need “negative sample”?

In conditional GAN's implementation, the inputs and labels of the discriminator are (fake_image, corresponding_condition, 0) and (real_image, corresponding_condition, 1). For example, (fake_dog, dog, ...
0
votes
2answers
145 views

What is the stop criteria of generative adversarial nets?

I have used the GANs (Generative Adversarial Networks) with a binary cross-entropy loss function (in both generator and discriminator). Throughout the training step, the variation of generator loss ...
0
votes
0answers
7 views

split neural network in two nets preserving weights in python [migrated]

In keras I would like to use the model with the initial layers of the structure for a given trained neuralnet with the weights i got for the training process. ...
32
votes
6answers
13k views

Can deep neural network approximate multiplication function without normalization?

Let say we want to do regression for simple f = x * y using standart deep neural network. I remember that there are reseraches that tells that NN with one hiden ...
4
votes
2answers
715 views

First-layer Visualizations in a neural network

I am reading the lectures on "Convolutional Neural Networks for Visual Recognition", and in this lecture they deal with first layer visualization. As you can see in the figure below- this figure ...
0
votes
0answers
13 views

Loss functions for segmentiation?

I am looking at segmenting images of cells for my neural network and found that touching cells have contrasting intensities and gradients. One paper used the MSE of gradients from a pixel to the GT to ...
0
votes
0answers
9 views

Large gap between train and validation losses

I try to train GAN to make super resolution, and in the beginning of my training (first iteration) there is a large gap between train and validation losses, and then next epoch it significally reduces....
6
votes
2answers
237 views

To what exactly does the term “activations” refer in neural networks?

Does it refer to the input or the output of the activation function? The literature seems to be inconsistent. A few examples: Activations = Input of the activation function Deep Learning Book, ...
0
votes
0answers
9 views

Training Loss Never Reached Convergence

I made a neural network with MLPCLassifier() module in scikit-learn. I set epoch until 100.000 to train the model, but why I never reach convergence?
0
votes
0answers
20 views

Predicting with Stateful LSTM

Knowing the nature of my time series problem, I am using a stateful LSTM to forecast one step ahead. My question is quite straightforward. Do I need to explicitly save and pass the hidden cell in ...

1
2 3 4 5
165