Before being fed to the neural network there are kernels applied to images for feature extraction.But, how do we understand that a particular kernel will help to extract the required feature for neural network.
There is absolutely no general answer to this question, no prinicipal method to determine these hyperparameters is known. A conventional approach is to look for similar problems and deep learning architectures which have already been shown to work. Than a suitable architecture can be developed by experimentation. However conventional kernel size's are 3x3, 5x5 and 7x7.
Otherwise, there are paper about this 1 and 2, you may want to take look to see the art of choosing hyper parameters in CNN.
Related
I am learning to use gabor filters to extract orientation and scale related features from images. On the other hand, Convolution Neural Network can also extract features including orientation and scale, is there any evidence that filters in CNN performs a similar function as gabor filters? Or pros and cons of both of them.
In my personal experience, in a traditional deep learning architecture (such as AlexNet) , when the layers near the beginning are visualized, they resemble gabor filters a lot.
Take this visualization of the first two layers of a pretrained AlexNet (taken from Andrej Karpathy's cs231n.github.io ). Some of the learnt filters look exactly like the Gabor Filters. So yes, there is evidence that CNN works (partly) the same way as Gabor Filters.
One possible explanation is that since the layers towards the beginning of a deep CNN are used to extract low level features (such as changes in texture), they perform the same functions as Gabor Filters. Features such as those detecting changes in frequency are so fundamental that they are present irrespective of the type of dataset the model is trained on. (Part of the reason why transfer learning is possible) .
But if you have more more data, you could possibly make a deep CNN learn much more high-level features than Gabor Filters, which might be more useful for the task you're extracting these features for (such as classification). I hope this provides some clarification.
I have designed a 3 layer neural network whose inputs are the concatenated features from a CNN and RNN. The weights learned by network take very small values. What is the reasonable explanation for this? and how to interpret the weight histograms and distributions in Tensorflow? Any good resource for it?
This is the weight distribution of the first hidden layer of a 3 layer neural network visualized using tensorboard. How to interpret this? all the weights are taking up zero value?
This is the weight distribution of the second hidden layer of a 3 layer neural:
how to interpret the weight histograms and distributions in Tensorflow?
Well, you probably didn't realize it, but you have just asked the 1 million dollar question in ML & AI...
Model interpretability is a hyper-active and hyper-hot area of current research (think of holy grail, or something), which has been brought forward lately not least due to the (often tremendous) success of deep learning models in various tasks; these models are currently only black boxes, and we naturally feel uncomfortable about it...
Any good resource for it?
Probably not exactly the kind of resources you were thinking of, and we are well off a SO-appropriate topic here, but since you asked...:
A recent (July 2017) article in Science provides a nice overview of the current status & research: How AI detectives are cracking open the black box of deep learning (no in-text links, but googling names & terms will pay off)
DARPA itself is currently running a program on Explainable Artificial Intelligence (XAI)
There was a workshop in NIPS 2016 on Interpretable Machine Learning for Complex Systems
On a more practical level:
The Layer-wise Relevance Propagation (LRP) toolbox for neural networks (paper, project page, code, TF Slim wrapper)
FairML: Auditing Black-Box Predictive Models, by Fast Forward Labs (blog post, paper, code)
A very recent (November 2017) paper by Geoff Hinton, Distilling a Neural Network Into a Soft Decision Tree, with an independent PyTorch implementation
SHAP: A Unified Approach to Interpreting Model Predictions (paper, authors' code)
These should be enough for starters, and to give you a general idea of the subject about which you asked...
UPDATE (Oct 2018): I have put up a much more detailed list of practical resources in my answer to the question Predictive Analytics - “Why” factor?
The weights learned by network take very small values. What is the reasonable explanation for this? How to interpret this? all the weights are taking up zero value?
Not all weights are zero, but many are. One reason is regularization (in combination with a large, i.e. wide layers, network) Regularization makes weights small (both L1 and L2). If your network is large, most weights are not needed, i.e., they can be set to zero and the model still performs well.
How to interpret the weight histograms and distributions in Tensorflow? Any good resource for it?
I am not so sure about weight distributions. There is some work that analysis them, but I am not aware of a general interpretation, e.g., for CNNs it is known that center weights of a filter/feature usually have larger magnitude than those in corners, see [Locality-Promoting Representation Learning, 2021, ICPR, https://arxiv.org/abs/1905.10661]
For CNNs you can also visualize weights directly, if you have large filters. For example, for (simpl)e networks you can see that weights first converge towards some kind of class average before overfitting starts. This is shown in Figure 2 of [The learning phases in NN: From Fitting the Majority to Fitting a Few, 2022, http://arxiv.org/abs/2202.08299]
Rather than going for weights, you can also look at what samples trigger the strongest activations for specific features. If you don't want to look at single features, there is also the possibility to visualize what the network actually remembers on the input, e.g., see [Explaining Neural Networks by Decoding Layer Activations, https://arxiv.org/abs/2005.13630].
These are just a few examples (Disclaimer I authored these works) - there is thousands of other works on explainability out there.
I'm trying to classify hotel image data using Convolutional neural network..
Below are some highlights:
Image preprocessing:
converting to gray-scale
resizing all images to same resolution
normalizing image data
finding pca components
Convolutional neural network:
Input- 32*32
convolution- 16 filters, 3*3 filter size
pooling- 2*2 filter size
dropout- dropping with 0.5 probability
fully connected- 256 units
dropout- dropping with 0.5 probability
output- 8 classes
Libraries used:
Lasagne
nolearn
But, I'm getting less accuracy on test data which is around 28% only.
Any possible reason for such less accuracy? Any suggested improvement?
Thanks in advance.
There are several possible reasons for low accuracy on test data, so without more information and a healthy amount of experimentation, it will be impossible to provide a concrete answer. Having said that, there are a few points worth mentioning:
As #lejlot mentioned in the comments, the PCA pre-processing step is suspicious. The fundamental CNN architecture is designed to require minimal pre-processing, and it's crucial that the basic structure of the image remains intact. This is because CNNs need to be able to find useful, spatially-local features.
For detecting complex objects from image data, it's likely that you'll benefit from more convolutional layers. Chances are, given the simple architecture you've described, that it simply doesn't possess the necessary expressiveness to handle the classification task.
Also, you mention you apply dropout after the convolutional layer. In general, the research I've seen indicates that dropout is not particularly effective on convolutional layers. I personally would recommend removing it to see if it has any impact. If you do wind up needing regularization on your convolutional layers, (which in my experience is often unnecessary since the shared kernels often already act as a powerful regularizer), you might consider stochastic pooling.
Among the most important tips I can give is to build a solid mechanism for measuring the quality of the model and then experiment. Try modifying the architecture and then tuning hyper-parameters to see what yields the best results. In particular, make sure to monitor training loss vs. validation loss so that you can identify when the model begins overfitting.
After 2012 Imagenet, all convolutional neural networks which performs good(state of the art) are adding more convolutional neural network, they even use zero padding to increase the convolutional neural network.
Increase the number of convolutional neural network.
Some says that dropout is not that effective on CNN, however it is not bad to use, but
You should lower the dropout value, you should try it(May be 0.2).
Data should be analysed. If it is low,
You should use data augmentation techniques.
If you have more data in one of the labels,
You are stuck with the imbalanced data problem. But you should not consider it for now.
You can
Fine-Tune from VGG-Net or some other CNN's should be considered.
Also, don't convert to grayscale, after image-to-array transformation, you should just divide 225.
I think that you learned CNN from some tutorial(MNIST) and you think that you should turn it to grayscale.
Does it make any sense to perform feature extraction on images using, e.g., OpenCV, then use Caffe for classification of those features?
I am asking this as opposed to the traditional way of passing the images directly to Caffe, and letting Caffe do the extraction and classification procedures.
Yes, it does make sense, but it may not be the first thing you want to try:
If you have already extracted hand-crafted features that are suitable for your domain, there is a good chance you'll get satisfactory results by using an easier-to-use machine learning tool (e.g. libsvm).
Caffe can be used in many different ways with your features. If they are low-level features (e.g. Histogram of Gradients), then several convolutional layers may be able to extract the appropriate mid-level features for your problem. You may also use caffe as an alternative non-linear classifier (instead of SVM). You have the freedom to try (too) many things, but my advice is to first try a machine learning method with a smaller meta-parameter space, especially if you're new to neural nets and caffe.
Caffe is a tool for training and evaluating deep neural networks. It is quite a versatile tool allowing for both deep convolutional nets as well as other architectures.
Of course it can be used to process pre-computed image features.
I am a newbie in machine learning and also in neural networks. Currently I'm taking a course at coursera.org about neural networks, but I don't understand everything. I have a little problem with my thesis. I should use a neural network, but I don't know how to choose the right neural network architecture for my problem.
I have a lot of data from web portals (typically online editions of newspapers, magazines). There is information about articles for example, name, text of article and release of article. There are also large amounts of sequence data that capture behavior of users.
My goal is to predict the popularity of an article (number of readers or clicks on article by unique user). I want to make vectors from this data and feed my neural network with these vectors.
I have two questions:
1. How do I create the right vector?
2. Which neural network architecture is best suited for this problem?
Those are very broad questions. You'll need to identify smaller issues if you want more exact answers.
How to create a right vector?
For text data, you usually use the vector space model. Best results are often obtained using tf-idf weighting.
Which neural network architecture is suitable for this problem?
This is very hard to say. I would start with a network with k input neurons (where k is the size of your vectors after applying tf-idf: you might also want to do some sort of feature selection to reduce the number of features. A good feature selection method is by using the chi squared test.)
Then, a standard network layout is given by using a single hidden layer with number of neurons equal to the average between the number of input neurons and output neurons. Then it looks like you only need a single output neuron that will output how popular the article is going to be (this can be a linear neuron or a sigmoid neuron).
For the neurons in your hidden layer, you can also experiment with linear and sigmoid neurons.
There are many other things you can try as well: weight decay, the momentum technique, networks with multiple layers, recurrent networks and so on. It's impossible to say what would work best for your given problem without a lot of experimentation.