Stack Exchange Network

Stack Exchange network consists of 174 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Visit Stack Exchange

Questions tagged [machine-learning]

Questions about machine learning functionality of Mathematica, for instance involving, but not restricted to, Predict, Classify, FindClusters and the various regression functions.

12
votes
1answer
553 views

Same neural network gives different results

I was playing with the neural network in Mathematica and found that the same neural network gives different classification results to the same image, which is very strange. Consider classifying this ...
5
votes
1answer
186 views

Response of Dynamic is slow

I wrote the following code to make an interactive GUI for Mathematica's Neural Network example of Digit Classification Here is the Code ...
12
votes
2answers
522 views

Implementing a Neural Network

TLDR: I am implementing a neural net in Mathematica and need help with back propagation. This is purely for the joy of implementing a neural network with a functional programming language. If someone ...
1
vote
1answer
64 views

Get a trained net and the training session properties?

In version 11.1 there is an additional signature for NetTrain: But this only returns the property, how can I get the net and the list of properties I want? Minimal example code: ...
4
votes
0answers
1k views

Neural network: Line detection in Image

I am playing around with Mathematica's new Neural Network toolbox and am trying to teach a network to detect lines in quite noisy images. An example Input image would look like the following: The ...
14
votes
1answer
249 views

WLNet format incompatible between version 11.0 and 11.1

It seems that there are some behavior changes of neural network layers from version 11.0 to 11.1 that leads to the incompatibility of WLNet format. Consider this example: Network exported in 11.0 <...
85
votes
1answer
10k views

Crack CAPTCHA using deep learning

Deep convolutional neural networks are very good at computer vision related tasks. Optical Character Recognition (OCR) is one important branch of computer vision. In fact, the convolution neural ...
8
votes
1answer
191 views

Adding computer vision datasets?

I find having ResourceData/Objects for machine learning datasets to be extremely useful, with their "TrainingData", "TestData" properties: Right now there's only a handful: And even fewer for ...
12
votes
1answer
522 views

Neural machine translation model using deep learning

In Mathematica version 11.1, SequenceAttentionLayer is introduced to model attentions in the recurrent neural network. It has been shown that attention mechanism ...
7
votes
1answer
122 views

Can Mathematica Neural Net make use of the REC file format for images like MXNet?

When using large set of images for training neural networks (>10000, size 250 x 250 each) the time to read the images is quite long. My first approach was to use the DataSet functionality of ...
5
votes
2answers
175 views

Inexplicable results from ClusteringComponents Agglomerate in version 11.1

I have data like this: data = {5, 12, 18, 19, 20, 23, 24, 26}; On Mathematica 11.1, clustering produces a variable final choice of clusters. For example: ...
17
votes
4answers
768 views

How to use Mathematica to train a network Using out of core classification?

I see there is doc about how to train a network Using out of core image classification and this question.But the object is only image. I want to use a binary file as data(Sequence to Sequence case),...
5
votes
1answer
286 views

How to build a NetChain by the information of ClassifierFunction?

NetChain provides a method NetExtract. I want to see how standClassify = Classify[trainingData, Method -> "NeuralNetwork"] works,beacause it has a better ...
5
votes
1answer
155 views

Using a classification Neural Net and parameter to ClassifierMeasurements[]

I created and trained a simple neural net, which takes some inputs and produces a 3-class classification. Then I feed it to CM[] function with some test data, and... nothing. Haven't been successful ...
42
votes
2answers
5k views

Object detection and localization using neural network

Four important computer vision tasks are classification, localization, object detection and instance segmentation (image taken from cs224d course): These four tasks are all built on top of the deep ...
12
votes
1answer
758 views

How to visualize attention?

In the articles about sequence attention we can see images like this: Here we see that while translating from French to English, the network attends sequentially to each input state, but sometimes it ...
10
votes
2answers
325 views

Mathematica 11.1 - Issue with Training Function for Sequence Data

I am attempting to use the new recurrent network functionality in Mathematica 11.1 Home Edition. I created a network that will attempt to predict future bytes in a raw audio file (using very small ...
16
votes
1answer
896 views

How to implement a stateful RNN?

If the model is stateless, the cell states are reset at each example. With the stateful model, all the states are propagated to the next example. Stateful RNNs can learn the long sequences. For ...
2
votes
0answers
98 views

Why my symbols been cleared?

I have about 60000 data to train. After minutes of calculation,all symbols are cleared. Is it a Bug? I have 8G memories ,but it's not enough.My cpu is i5-4590. Here are my codes: ...
6
votes
3answers
534 views

how to take ImageFeatures?

I want to know how to transform images to feature vector. There is a function FeatureExtractor.And it has some methods. ...
5
votes
0answers
155 views

GPU Acceleration and the Predict function [duplicate]

With the Mathematica Predict function, does it automatically use GPU acceleration? For example, with NetTrain, one has to specify TargetDevice->"GPU" to show they want to perform the calculation on ...
1
vote
0answers
143 views

Trouble loading MXNetLink in console mode [closed]

I'm using v11.1 on ubuntu, any ideas what is wrong here: In[3]:= << MXNetLink` ...
3
votes
0answers
89 views

How can I process many image files with Classify?

Classify in Mathematica is a highly automated machine learning tool. Though it is not good for production purposes, sometimes it can used as a benchmark tool. ...
4
votes
0answers
187 views

Is Mathematica a good choice for a scientific project with Machine Learning?

I'm a chemist with some rudimentary programming skills and in the middle of the year I'll be starting a project concerning machine learning, so, I'm sorry if I'm going to ask two questions in this ...
5
votes
1answer
207 views

How to write a RCNN (region-cnn)?

Has anyone end-to-end trained a single net that can find the location of objects in an image (with v11.1)? Here's an example of a region-cnn in action, and this is a good description of the approach.
3
votes
0answers
227 views

radial basis function network in Mathematica 11

I wanted to compare some (pre-existing) Python code I have for Radial basis function (neural) networks (RBFNN) for the goal of function approximation, with what I could obtain with Mathematica 11.0. ...
5
votes
0answers
105 views

Optimize a network topology?

I'd like to know if one can use optimization function like Maximize[] or BayesianMaximize[] to optimize the hyper-parameters of ...
6
votes
1answer
294 views

Implementing Resnet as a NetGraph?

Does anyone know of a Mathematica v11.1 implementation for any of the Resnets, e.g. a NetModel or WLNet for Resnet-1000? Loading the MXNet weights from the model zoo for both resnet or resnext doesn'...
5
votes
0answers
332 views

How to train a visual attention model?

In version 11.1 there is a new layer type SequenceAttentionLayer[]. I'm interested in how might this be used to train a recurrent visual attention model (see the ...
37
votes
4answers
2k views

Generative Adversarial Network

Generative adversarial networks (GAN) is regarded as one of "the most interesting idea in the last ten years in machine learning" by Yann LeCun. It can be used to generate photo-realistic images that ...
3
votes
1answer
84 views

What is the point of ActiveClassification?

I must be missing something here - it seems that the first argument to ActiveClassification is your target function, but if you already know that function then why ...
6
votes
1answer
120 views

Is it possible to convert the result of NetTrain to a PredictorFunction?

I would like to use PredictorMeasurements on the result of NetTrain, but it says it expects a ...
3
votes
0answers
91 views

Is it possible to train a FindObject and mark the object?

We can use Classify to train a classifier to classify an object. Can we make the classify more intelligent or work like FindFaces...
1
vote
1answer
161 views

Constructing List of Rules from imported data

I would like to create a List of Rules data structure to be used in the Classify function. I have sample input data stored in ...
3
votes
1answer
236 views

Custom NetEncoder for lower resolution image input in a pre-trained Inception network

I have come across this neat explanation on how to adjust a pre-trained Inception neural net to an arbitrary number of classes by truncating the top layers. However, my input data does not come in $...
9
votes
2answers
220 views

Options for ValidationSet accuracy from NetTrain?

With NetTrain[] if you use the option ValidationSet->Scale[.1] it will run a validation test on 10% of the training data ...
3
votes
1answer
115 views

Dimensions of ConfusionMatrix

Maybe a simple question and I have overseen something, but: Why is the ConfusionMatrix in ClassifierMeasurements not a quadratic matrix? From the documentation: "counts c_{i,j} of class i examples ...
24
votes
2answers
665 views

How to make a custom NN layer in Mathematica?

Is there a way to create our own custom layers for Neural Networks in Mathematica ? I suspect it should be possible, any pointers or directions on how to go about it? I am interested in ...
2
votes
2answers
263 views

Concise / idiomatic replacement for a For loop [closed]

I am using the following For loop to measure the misclassification errors in a Classifier. ...
1
vote
1answer
52 views

ClassifierFunction::mlbddataev: The data being evaluated is not formatted correctly

I am trying to run the NearestNeighbor classifier on the wisc_bc_data.csv dataset as follows: ...
14
votes
1answer
682 views

Implementing t-SNE (t-Distributed Stochastic Neighbor Embedding)

Context Six months ago @M.R. asked about an implementation of the t-Distributed Stochastic Neighbor Embedding (t-SNE) algorithm by van der Maaten and Hinton (2008). (@M.R.'s question) @Alexey ...
2
votes
0answers
161 views

Sentimental Analysis Classification Problem [closed]

I am keen on working with Mathematica for my sentimental analysis project. I have just started and been working around with some examples. During which I found: ...
2
votes
0answers
57 views

NetTrain with OpenCL [duplicate]

When working with NetTrain I get the following error: TargetDevice -> {GPU,0} could not be used, please ensure that you have a compatible graphics card and have installed CUDA drivers. Is it ...
5
votes
2answers
135 views

SoftmaxLayer Incompatible types

Trying to create U-Net similar to this paper from this GitHub project but keep getting "incompatible type" error on the final softmax layer. Any ideas for how to organize data to satisfy Softmax layer?...
3
votes
2answers
214 views

Mathematica NetChain With Mixed Inputs (Numerical and Categorical)

I have been experimenting with Mathematica Tensor Flow (NetChain and NetTrain) and I would love to know if a simple step exists to present an input with a mixture of numerical and categorical ...
2
votes
1answer
399 views

CNN Nettrain problem

I 500 pictures of tables and chairs each (849x849) and want to train a convolutional neural net to classify new pictures accordingly. This is my code so far: ...
2
votes
0answers
127 views

Why not all properties in ClassifierMeasurements are available?

I'm doing an experiment of objects classification on dataset CIFAR-10 and I'm doing this using a neural network. The basic code of this experiment is come from the demo on Mathematica website and I ...
5
votes
1answer
196 views

Predict on Raspberry Pi is incredibly slow or simply hangs

I'm testing the latest Mathematica available on the Raspberry Pi (10.3.1 for Linux ARM (32-bit) (January 11, 2016)) for machine learning but I'm having issues with the default behaviour of Predict. ...
27
votes
1answer
2k views

How to export an MXNet?

I was hoping there was some way to generate the .params and .json file needed to define an MXNet model, from a network trained using NetTrain[] in Mathematica. I ...
12
votes
1answer
336 views

How to determine which Layers should be used in my NetChain?

I have a answer in this post which is based on neural network,but I have to say that reault is too bad and have no practicability.I think the reason is that selection of network layers.So I give some ...