Questions tagged [deep-learning]

a new area of Machine Learning research concerned with the technologies used for learning hierarchical representations of data, mainly done with deep neural networks (i.e. networks with two or more hidden layers), but also with some sort of Probabilistic Graphical Models.

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

“Memory Error” - deploying in AWS elastic beanstalk free tier

While trying to deploy flask image classification model on elastic beanstalk, I am getting this memory error. Is it because of the limited size of 512 MB provided for uploading source code? I face ...
1
vote
1answer
15 views

Predicting Multiple Values Values Using Time Series Forecasting

I want to illustrate my question with the following example: I have a wholesale company through which I sell 200 products: P1,P2,P3 .... P200 to a 1000 customers ...
0
votes
0answers
4 views

How to store and query biometric data for an authentication system?

I am trying to design, and hopefully implement, an authentication system which centers around the use of biometric images. I plan to use different machine learning and deep learning techniques to help ...
0
votes
0answers
6 views

Estimating coefficients of AR(n) process with deep learning

Can a deep learning model estimate parameters of an AR(n) process given a time series generated by such process? What kind of deep-learning model would you recommend? Note: I am very new to machine ...
4
votes
1answer
22 views

Is there an inherent recency bias in deep learning?

When working with very large models within Deep Learning, training often takes long and requires small batch sizes due to memory restrictions. Usually, we are left with a model checkpoint after ...
0
votes
0answers
4 views

Problem with weights having the same shape[0] during forward propagation and preventing the dot product from working

I'm trying to implement forward propagation in my neural network but it doesn't seem to work and I suspect it is because the number of neurones from one layer to the next are not the same. ...
1
vote
1answer
19 views

Dying leaky ReLU

I am trying to train a deep neural network but I am having dying ReLU problem. I am using leaky Relu but still have the same problem. Isn't leaky relu supposed to not have such problems?
0
votes
0answers
5 views

Is class discrimination is possible in Class Activation Mapping(CAM)?

I worked with Gradient Weighted class activation mappings(Grad-CAM) to understand and implement interpretability in Deep neural networks. I can also switch to a particular class(by selecting the ...
-1
votes
0answers
18 views

How to train my model using keras GPU?

I want to use google colab to train my model, I changed the execution type to GPU! is there somthing I should be adding to the code so that it trains on GPU I am using keras?
0
votes
0answers
10 views

Preprocess problem Faster RCNN in tensorflow object detection API

I've wrapped meta-architecture with the code below: ...
1
vote
1answer
20 views

How to train a neural network on multiple objectives?

I have a multi-class neural network classifier that has K classes(products). For every row, only one of the classes will be 1 at a time. Now, this approach works fine if I have only 1 objective to ...
0
votes
0answers
12 views

Best Practices for Bayesian Deep Learning

Does anyone have a set of best practices for Bayesian Deep Learning? I'm working on a model with a CNN block followed by a dense interpretation block. I am looking to replace the dense interpretation ...
0
votes
1answer
32 views

How can I fixed the filter and Kernel Size of a CNN?

I have created 4 x 4 2d images from a signal. Now, I want to feed this data to a Convolutional neural Network. How I can choose the nubmber of ...
0
votes
1answer
15 views

Sneakers representation learning

I am trying to make a model which would take an image of shoes as an input and output a meaningful N-dimensional embedding of the shoes, so that they could be searchable/comparable/clustered and used ...
1
vote
0answers
31 views

For short sentences(max length 10 ), which Name entity recognition algorithm is good?

My Training data look like this . I have to recognize 4 class for each sentence. Any algorithm , which have some learning parameters Means not rule based approach . So which method is good for my ...
0
votes
1answer
20 views

Why this TensorFlow Transformer model has Linear output instead of Softmax?

I am checking this official TensorFlow tutorial on a Transformer model for Portuguese-English translation. I am quite surprised that when the Transformer is created, their final output is a Dense ...
0
votes
0answers
11 views

Training on compressed video, testing on uncompressed images, performance degradation?

In my application, I have to collect training data from a single camera. At test-time, the camera frames will be fed live to the network, without being saved in a lossy format in between. Now I wonder ...
0
votes
1answer
22 views

PyTorch cross_entropy with 3D data (RNN/LSTM)

I am working on LSTMs and I want to compute cross_entropy loss, given X and y. X.shape: (batch_size, time_steps, number_of_classes) y.shape: (batch_size, time_steps) y contains the ground truth ...
-1
votes
1answer
24 views

Can we identify that an academic dataset was used for commercial purpose [closed]

There are many datasets released on the internet. Authors of many of these datasets state that the datasets are strictly for academic usage and not for commercial purposes. Although some datasets are ...
0
votes
0answers
7 views

How to train an LSTM model with data that has multiple input rows per day but only one row of label/ground-truth (output) data per day

I am doing a sleep data science experiment and I need a model that outputs multiple columns sleep quality measurement values (that are decimal numbers) for each input. For training, I collected data ...
0
votes
0answers
17 views

What sort of models work for unsupervised reinforcement learning, or is deep learning the way?

I'm setting out on an adventure to automate the statuses of the lights around my home. The lights should have different brightness in the range [0, 100] depending on some factors, which I have boiled ...
2
votes
2answers
33 views

How to compute score and predict for outcome after N days

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). I did post here ...
0
votes
0answers
9 views

Single Image/Input Training [closed]

I have an Object Detection model that also classifies the objects in the image. In my experiment, I am using a Single Shot Detector model with bounding boxes. The model has already been trained with a ...
1
vote
2answers
20 views

Can I create a layer with multiple rnn cell ? [question about a paper]

I am trying to implement https://dl.acm.org/doi/pdf/10.1145/3269206.3271794 . Structure: As it said: In particular, we integrate the embedding vectors learned from each individual recurrent encoder ...
0
votes
0answers
14 views

Tensorflow Conv2d Input problem [closed]

I have an "image" with data as: points: x, y columns features: red, green, blue, weight(for pixel) columns, corresponding to each point They are saved as tensors with 100 length in each ...
0
votes
0answers
6 views

Can we do autoregressive using pad_packed_sequence?

I’m curious about if we want to do the autoregressive manner. Is it possible to do with implemented using pad_packed_sequence and pack_padded_sequence input to some recurrent network? Because we need ...
1
vote
0answers
12 views

How do I create a dok matrix with split files correctly?

I let my model run normally and have defined an early stopping as a callback. The model breaks off, I let it run through without early stopping, loss and val_loss go further and further apart (see ...
0
votes
0answers
7 views

torch.save(the_model, PATH) vs torch.save(the_model.state_dict(), PATH) - model loading incorrectly for one method

I just now noticed that the model does not get loaded correctly if I use the the_model.state_dict() method to save it. On the other hand, using ...
0
votes
0answers
5 views

How to generate fixed number of superpixels?

A lot of work regarding Graph Neural Networks require fixed number of nodes. In the case of image processing using graph, the image representation is often super pixels (like in this work https://...
-1
votes
0answers
13 views

Is Transformer better then GRU for human acitivity recognition

I am working on human activity recognition and I was wondering If there are documented studies( Papers) about GRU vs Transformers in this context?
0
votes
0answers
10 views

Usage of Doc2Vec as feature extractor for text classification of websites with political articles

I have gathered political articles from polish websites for my engineering thesis. The main goal is to try to predict the website that input text belongs to. So for this few websites I want to create ...
4
votes
4answers
602 views

What is the difference between AI, ML, NN and DL? [closed]

What is the difference between the following four categories: Artificial Intelligence (AI) Machine Learning (ML) Neural Network (NN) Deep Learning (DL) Data Science My current understanding is that ...
0
votes
0answers
8 views

Help understanding input to biaxial network for generating music

I am reading Composing Music With Recurrent Neural Networks by Daniel D. Johnson. But I am really confused about the input passed to this network. If we pass notes of music along the time axis, then ...
0
votes
0answers
19 views

Is padding the right way to allow your model to make prediction with test sequences of shorter lengths?

Say I have a RNN-lstm encoder-decoder model trained on fixed timesteps (no padding when training, all sequences are treated as if having the same lengths). My testing criteria requires me to provide ...
0
votes
0answers
8 views

Working Behavior of BERT vs Transformers vs Self-Attention+LSTM vs Attention+LSTM on the scientific STEM data classification task?

So I just used BERT pre-trained with Focal Loss to classify Physics, Chemistry, Biology and Mathematics and got a good f-1 macro of 0.91. It is good given it only had to look for the tokens like ...
0
votes
0answers
30 views

Stateful LSTM in Deployment

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 ...
0
votes
1answer
20 views

What is the difference between GPT blocks and Transformer Decoder blocks?

I know GPT is a Transformer-based Neural Network, composed of several blocks. These blocks are based on the original Transformer's Decoder blocks, but are they exactly the same? In the original ...
0
votes
1answer
28 views

my model is predicted new images wrong?

I used a CNN network unet for a segmentation task this is the architecture I used ...
1
vote
1answer
24 views

ValueError: No gradients provided for any variable

I have this error when running training on my model. I found this issue on different sites, but could not find a solution to my problem. Here is my model : ...
0
votes
0answers
14 views

How to customize what happen in fit method in tensorflow 2.0

I'm studing how to customize what happen in fit method with tensorflow 2.0 and I'm following this link: https://www.tensorflow.org/guide/keras/customizing_what_happens_in_fit, but I noticed some ...
0
votes
1answer
21 views

what is the largest network used for image recognition/segmentation?

What is the largest network (in number of params and layers) considered in the literature for image recognition/segmentation task? I am in particular interested in ResNet architectures. Any ...
1
vote
1answer
34 views

How to normalize technical skills of IT using machine learning?

I have a huge collection of skills collected/scraped from various online sources. It was huge effort done by our team. Now, the biggest challenge we are facing is trying to normalize the skills back ...
0
votes
1answer
18 views

multi regression for energy data

I'm trying to develop a multi regression model to predict energy consumption during one day period. X-set dimension is (10178, 52) and consist of 52-feature and Y-set dimension is (10178, 48) as ...
3
votes
1answer
119 views

Which learning rate should I choose?

I'm training a segmentation model, Unet++, on 2d images and I am now trying to find the optimal learning rate. The backbone of the model is Resnet34, I use Adam optimizer and the loss function is the ...
1
vote
0answers
20 views

YOLOv1 algorithm - how to determine predictor responsibility

I am researching Yolo detector, and have read the original paper, but still have some confusion and a few questions regarding the assignment of bounding box predictors to ground truth at training time ...
0
votes
2answers
29 views

What does num_negatives mean? [closed]

Does anyone know what does the num_negatives mean? What is it for? ...
1
vote
1answer
20 views

Neural networks with not-fixed dimension for input and output

I would like to know if it exists a model/method which can deal with input and output of different dimension. For example, let us say that the maximum number of info we could have is 6 features and 5 ...
0
votes
1answer
33 views

Augmentation on test dataset and validation dataset

I'm training a segmentation model (computer-vision). Thus, my dataset contains images and masks (binary segmentation of objects). I'm augmenting the training dataset (applying random crop, rotation or ...
0
votes
0answers
20 views

How should I encode 'dynamic' features (with multiple instances) along with 'static' features (single instances)?

Suppose I have to predict if a certain product from an assembly line in a factory will be a scrap. This product has let's say 'static' data like a certain shape. A certain vendor, etc. And, it can ...
0
votes
1answer
35 views

How should I sample my validation set if I randomly sample training data?

I have: training dataset of size 150k. validation dataset of size 19k. At each epoch I randomly sample without replacement 10k datapoints for training because I get Out of Mem Errors. I need to ...

1
2 3 4 5
70