What's the deal with Deno? We talk with a major contributor to find out. Listen now.

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
10
votes
1answer
4k views

What is a tower?

In many tensorflow tutorials (example) "towers" are mentioned without a definition. What is meant by that?
8
votes
4answers
13k views

Pattern Recognition on Financial Market

Which machine learning or deep learning model(has to be supervised learning) will be best suited for recognizing patterns in financial markets ? What I mean by pattern recognition in financial market ...
1
vote
1answer
325 views

Digits Localization on Streets View House Numbersm

I am trying to learn a bit of deep learning playing with the Street View House Numbers data set. I have managed to recognize sequences of digits and I'd like now to train a CNN to localize digits and ...
6
votes
1answer
2k views

Multi scale CNN Network Python

I created a multi-scale CNN in python keras. The network architecture is similar to the diagram. Here, same image is fed to 3 CNN's with different architectures. The weights are NOT shared. I ...
1
vote
1answer
2k views

Recognize Street View House Numbers

I am new to deep learning and I am trying to train a NN to recognize house numbers gathered from street view. I have already managed to recognized MNIST sequence of hand written digits by means of a ...
1
vote
0answers
66 views

Implementation and interpretation of recursive convolution neural net(RCNN)

I am working on to implement the approach in the paper https://arxiv.org/pdf/1603.03101v1.pdf on detecting text in wild using recursive convolution neural net and attention modelling. Being a bit new ...
5
votes
1answer
1k views

RELU vs Pooling

Does RELU means to change pixel value to 0 if it is negative anywhere , and later if we apply maximum pooling then what is the use of RELU because in this step we choose maximum value so no matter it ...
1
vote
1answer
199 views

Extending a trained neural network for a larger input

I have a seq2seq conversational model (based on this implementation) trained on the Cornell movie dialogs. Now I want to fine-tune it on a much smaller dataset. The new data comes with the new words, ...
1
vote
0answers
1k views

Neural net learning only one class?

I have setup a four layer CNN designed to predict two classes. The two classes are more or less in the same ratio. The negative class is 55% of the data and the positive class makes up the remaining. ...
3
votes
1answer
722 views

Predict a tree structure out of nodes with different features

The Problem Suppose we have a representation of a document's text layout as in the image below: Here, each rectangle represents a chunk of text (one word, an expression or even part of a sentence). ...
1
vote
0answers
58 views

Spark Deeplearning4j Training Problem

I am training a model in Spark using Dl4J library in Yarn-Cluster mode. When I train the model on 2 lakh data (approx 200MB) then the job succeeds but when I go to train the model with 3 lakh data (...
6
votes
3answers
12k views

Multiple output classes in keras

I'm trying to predict movie genres using a neural network. I initially considered using a softmax layer as my output layer, but since a movie can have multiple genre labels, how should my output be? ...
4
votes
3answers
1k views

Handwritting Recognition moving from character level to word level

Given the experience on MIST, I try this problem as a character level. I have a handwritten text and I want to "OCR" it. Even though I made progresses with openCV (on the image pre-processing, ...
1
vote
2answers
465 views

Is the graphic of deep residual networks wrong?

I am currently wondering if the following graphic of deep residual networks is wrong: I would say the graphic describes $$\varphi \left (W_2 \varphi(W_1 x) + x \right ) \qquad \text{ with } \varphi =...
1
vote
1answer
334 views

the feasibility of image processing techniques for physics based images

When building deep learning models for image analytics-related applications, we sometimes apply various types of operations to enhance the image, such as an image denoising operation. In my study, we ...
0
votes
2answers
1k views

Mapping sequences of different lengths to fixed vector - Python

I am trying to make a chatbot using a deep neural network in python using keras. The problem I am having is that for the deep neural network to work, the input dimension has to be fixed. So my ...
0
votes
1answer
124 views

Computing weights in batch gradient descent

I have a reasonably large set of images that I want to classify using a neural network. I can't fil them all into memory at once, so I decided to process them in batches of 200. I'm using an cross-...
51
votes
3answers
17k views

What is the difference between “equivariant to translation” and “invariant to translation”

I'm having trouble understanding the difference between equivariant to translation and invariant to translation. In the book Deep Learning. MIT Press, 2016 (I. Goodfellow, A. Courville, and Y. Bengio)...
4
votes
5answers
2k views

Can we train a neural network to tell if an object is present or not in an Image?

I am new to machine learning, working on object detection, but not interested in the location of the object in the image, so I just want to know is it possible to train such a neural network, if yes, ...
7
votes
2answers
11k views

When is something a Deep Neural Network (DNN) and not NN?

When would a neural network be defined as a Deep Neural Network (DNN) and not a NN? A DNN as I understand them are neural networks with many layers, and simple neural networks usually have fewer ...
27
votes
6answers
10k views

Why do convolutional neural networks work?

I have often heard people saying that why convolutional neural networks are still poorly understood. Is it known why convolutional neural networks always end up learning increasingly sophisticated ...
3
votes
0answers
302 views

TensorFlow doesn't learn when input=output (or probably I am missing something)

I have used TensorFlow before for training a model to recognize images (something similar to this example: https://github.com/tensorflow/models/tree/master/inception). I was trying to do something ...
2
votes
1answer
931 views

Clipping threshold of softmax layer

I recently came across a paper on using (rather simple version of) LSTM for sentiment classification, and it describes its network settings as: We randomize the parameters with uniform distribution ...
5
votes
2answers
4k views

What is the best deep learning library for scala? [closed]

Does any one has a recommendation for what libraries to use for deep learning?
6
votes
1answer
6k views

Possible to use different learning rate for different neuron in Keras/Tensorflow?

The simplest example is to have faster/slower learning rates in the upper/lower layers of a network. I found this post on tensorflow. Is there a similar trick in Keras? Going one step further, can ...
1
vote
0answers
652 views

Does the time to train a model using keras increase linear with epoches?

Training a model in keras can go very fast. I trained the following model. To train the model with 1 epoch finishes in a few second. However, using 20 epochs it lasts half an hour or so? An epoch ...
2
votes
1answer
77 views

What knowledge should I gain for developing a supervised image processing software that learns how to edit photos based on past behavior? [closed]

I have done several machine learning projects but all of them have been connected to the traditional machine learning (predictions, classifications, etc.). I have currently been offered a project to ...
2
votes
3answers
727 views

How do I use Convolutional Neural Nets to classify if there are rain drops or soil on the camera lens?

I have certain videos for which the frames are labeled either as dirty (meaning the camera lens is occluded by soil or rain) or as clean. The goal is to test a convolutional neural net on this data to ...
5
votes
1answer
247 views

How are per-layer-detected-patterns in a trained CNN plotted?

In the case my question is not clear, I am talking about the patterns that are detected in each of the layers of an image-trained Convolutional Neural Network (CNN). Take the following image as an ...
1
vote
3answers
3k views

My Neural Network is not learning anything

I try to train convolutional neural net for a classification problem. However, my neural net is not learning anything. It guesses only two labels and ignores the rest. Even though I try to train to ...
2
votes
1answer
1k views

How hidden layer is made binary in Restricted Boltzmann Machine (RBM)?

In RBM, in the positive phase for updating the hidden layer(which should also be binary), [Acually consider a node of h1 ∈ H(hidden layer vector)] to make h1 a binary number we compute the probability ...
17
votes
3answers
10k views

How to add a new category to a deep learning model?

Say I have done transfer learning on a pre-trained network to recognize 10 objects. How do add a 11th item that the network can classify without losing all the 10 categories that I already trained nor ...
1
vote
1answer
342 views

# of iterations in Restricted Boltzmann Machine (RBM)

I have a training set, I provide it (consider a data from training set) to the visible layer. Then the normal process is followed, i.e. Positive Phase-> Negative Phase-> Reconstruction of weights, ...
3
votes
3answers
10k views

Training my neural network to overfit my training dataset

I want to train my neural net to overfit the training data. Should I just keep fitting my model with the same training data by using K-fold validation and setting epochs to infinity? Then, after it ...
4
votes
5answers
868 views

Deep Learning Project to Predict Stock Prices

So I have a background in computer programming and a little in machine learning in general. What I would like to do is create a fun project in A.I. with deep learning. I have a dataset that has a ...
1
vote
1answer
322 views

How is Amazon using Deep Learning in the new Amazon Go?

In the Amazon Go page they claim to use "deep learning". Where is deep learning used in the "Go" service? Is it computer vision or inventory management?
6
votes
3answers
6k views

Crop background from Image

I try to write a program to crop background from an image. This is a sample of my training data. I have images with and without a background. (manually cropped) The background is always similar (...
2
votes
0answers
989 views

Is finetuning from a pretrained model always better than training from scratch?

At the worst case scenario, we could treat the pretrained weights as a random initialization, same as what we would do for training from scratch, right? If that is the case, then wouldn't it be better ...
1
vote
0answers
741 views

Image as input and output in keras

I am trying to make a model of this image. Here is the relevant code: ...
11
votes
1answer
2k views

Why TensorFlow can't fit simple linear model if I am minimizing absolute mean error instead of the mean squared error?

In Introduction I have just changed loss = tf.reduce_mean(tf.square(y - y_data)) to ...
2
votes
1answer
1k views

How is dimensionality reduction achieved in Deep Belief Networks with Restricted Boltzmann Machines?

In neural networks and old classification methods, we usually construct an objective function to achieve dimensionality reduction. But Deep Belief Networks (DBN) with Restricted Boltzmann Machines (...
0
votes
1answer
319 views

How many vectors does paragraph vector generate for each paragraph?

For example,if I have a corpus with two paragraphs, does paragraph vector generate two vectors?Additionally, on Distributed Representations of Sentences and Documents (Q. Le, T. Mikolov) paper I do ...
3
votes
1answer
3k views

What's the best way to tune the regularization parameter in neural nets

I'm tuning the regularization parameter of a neural net (L2 regularization) using a grid. Starting with values 0.0005, 0.005, 0.05, 0.5, 5. Then if ...
3
votes
1answer
2k views

Autoencoder and Neural Network Overfitting in terms of parameter number?

I have 1100 sequences for 2 classes. Of them 400 are from one class 1 and ...
2
votes
1answer
1k views

Train a RNN with strings using positive and negative samples

I'm trying to recognize strings coming from a regular language, using positive and negative samples by Recurrent Neural Networks. In particular, I tried to use the rnnlib by Alex Graves, but I had ...
2
votes
1answer
193 views

Backpropagation derivation problem

I read a few tutorials on neural network backpropagation and decided to implement one from scratch. I tried to find this single error for the past few days I have in my code with no success. I ...
7
votes
1answer
1k views

After the training phase, is it better to run neural networks on a GPU or CPU?

My understanding is that GPUs are more efficient for running neural nets, but someone recently suggested to me that GPUs are only needed for the training phase. Once trained, it's actually more ...
1
vote
0answers
77 views

Neural Network: how to utilize weakly-/unsupervised data to improve supervised network?

Let's consider one has built a fully-supervised neural network for some task, e.g. localizing an object in various scenes. As you can imagine, it is quite time-consuming to label data: one has to ...
0
votes
1answer
122 views

Does the performance of neural networks depend on the method used to unroll weights ?

Lets say we have weights(theta1 and theta2) of neural net as: theta1 =[1, 2, 3] theta2= [4, 5, 6] If we unroll these weights into a single dimension array in ...
123
votes
6answers
167k views

How to draw Deep learning network architecture diagrams?

I have built my model. Now I want to draw the network architecture diagram for my research paper. Example is shown below:

1
63 64
65
66 67
69