This seems to be a fundamental question which some of you out there must have an opinion on. I have an image classifier implemented in CNTK with 48 classes. If the image does not match any of the 48 classes very well, then I'd like to be able to conclude that it was not among these 48 image types. My original idea was simply that if the highest output of the final Softmax layer was low, I would be able to conclude that the test image matched none well. While I occasionally see this occur, in most testing, Softmax still produces a very high (and mistaken) result when handed an 'unknown image type'. But maybe my network is 'over fit' and if it wasn't, my original idea would work fine. What do you think? Any way to define a 49-th class called 'none-of-the-above'?
You really have these two options indeed--thresholding the posterior probabilities (softmax values), and adding a garbage class.
In my area (speech), both approaches are their place:
If "none of the above" inputs are of the same nature as the "above" (e.g. non-grammatical inputs), thresholding works fine. Note that the posterior probability for a class is equal to one minus an estimate of the error rate for choosing this class. Rejecting anything with posterior < 50% would be rejecting all cases where you are more likely wrong than right. As long as your none-of-the-above classes are of similar nature, the estimate may be accurate enough to make this correct for them as well.
If "none of the above" inputs are of similar nature but your number of classes is very small (e.g. 10 digits), or if the inputs are of a totally different nature (e.g. a sound of a door slam or someone coughing), thresholding typically fails. Then, one would train a "garbage model." In our experience, it is OK to include the training data for the correct classes. Now the none-of-the-above class may match a correct class as well. But that's OK as long as the none-of-the-above class is not overtrained--its distribution will be much flatter, and thus even if it matches a known class, it will match it with a lower score and thus not win against the actual known class' softmax output.
In the end, I would use both. Definitely use a threshold (to catch the cases that the system can rule out) and use a garbage model, which I would just train it on whatever you have. I would expect that including the correct examples in training will not harm, even if it is the only data you have (please check the paper Anton posted for whether that applies to image as well). It may also make sense to try to synthesize data, e.g. by randomly combining patches from different images.
I agree with you that this is a key question, but I am not aware of much work in that area either.
There's one recent paper by Zhang and LeCun, that addresses the question for image classification in particular. They use large quantities of unlabelled data to create an additional "none of the above" class. The catch though is that, in some cases, their unlabelled data is not completely unlabelled, and they have means of removing "unlabelled" images that are actually in one of their labelled classes. Having said that, the authors report that apart from solving the "none of the above" problem, they even see performance gains even on their test sets.
As for fitting something post-hoc, just by looking at the outputs of the softmax, I can't provide any pointers.
Related
I am coding a spell-casting system where you draw a symbol with your wand (mouse), and it can recognize said symbol.
There are two methods I believe might work; neural networking and an "invisible grid system"
The problem with the neural networking system is that It would be (likely) suboptimal in Roblox Luau, and not be able to match the performance nor speed I wish for. (Although, I may just be lacking in neural networking knowledge. Please let me know whether I should continue to try implementing it this way)
For the invisible grid system, I thought of converting the drawing into 1s and 0s (1 = drawn, 0 = blank), then seeing if it is similar to one of the symbols. I create the symbols by making a dictionary like:
local Symbol = { -- "Answer Key" shape, looks like a tilted square
00100,
01010,
10001,
01010,
00100,
}
The problem is that user error will likely cause it to be inaccurate, like this "spell"'s blue boxes, showing user error/inaccuracy. I'm also sure that if I have multiple Symbols, comparing every value in every symbol will surely not be quick.
Do you know an algorithm that could help me do this? Or just some alternative way of doing this I am missing? Thank you for reading my post.
I'm sorry if the format on this is incorrect, this is my first stack-overflow post. I will gladly delete this post if it doesn't abide to one of the rules. ( Let me know if there are any tags I should add )
One possible approach to solving this problem is to use a template matching algorithm. In this approach, you would create a "template" for each symbol that you want to recognize, which would be a grid of 1s and 0s similar to what you described in your question. Then, when the user draws a symbol, you would convert their drawing into a grid of 1s and 0s in the same way.
Next, you would compare the user's drawing to each of the templates using a similarity metric, such as the sum of absolute differences (SAD) or normalized cross-correlation (NCC). The template with the lowest SAD or highest NCC value would be considered the "best match" for the user's drawing, and therefore the recognized symbol.
There are a few advantages to using this approach:
It is relatively simple to implement, compared to a neural network.
It is fast, since you only need to compare the user's drawing to a small number of templates.
It can tolerate some user error, since the templates can be designed to be tolerant of slight variations in the user's drawing.
There are also some potential disadvantages to consider:
It may not be as accurate as a neural network, especially for complex or highly variable symbols.
The templates must be carefully designed to be representative of the expected variations in the user's drawings, which can be time-consuming.
Overall, whether this approach is suitable for your use case will depend on the specific requirements of your spell-casting system, including the number and complexity of the symbols you want to recognize, the accuracy and speed you need, and the resources (e.g. time, compute power) that are available to you.
I have estimated a complex hierarchical model with many random effects, but don't really know what the best approach is to checking for convergend. I have complex longitudinal data from a few hundred individuals and estimate quite a few parameters for every individual. Because of that, I have way to many traceplots to inspect visually. Or should I really spend a day going through all the traceplots? What would be a better way to check for convergence? Do I have to calculate Gelman and Rubin's Rhat for every parameter on the person level? And when can I conclude that the model converged? When absolutely all of the thousends of parameters reached convergence? Is it even sensible to expect that? Or is there something like "overall convergence"? And what does it mean when some person-level parameters did not converge? Does it make sense to use autorun.jags from the R2jags package with such a model or will it just run for ever? I know, these are a lot of question, but I just don't know how to approach that.
The measure I am using for convergence is a potential scale reduction factor (psrf)* using the gelman.diag function from the R package coda.
But nevertheless, I am also quickly visually inspecting all the traceplots, even though I also have tens/hundreds of them. It can be really fast if you put them in PNG files and then quickly go through them using e.g. IrfanView (let me know if you need me to expand on this).
The reason you should inspect the traceplots is pretty well described by an example from Marc Kery (author of great Bayesian books): see "Never blindly trust Rhat for convergence in a Bayesian analysis", here I include a self explanatory image from this email:
This is related to Rhat statistics while I use psrf, but it's pretty likely that psrf suffers from this too... and better to check the chains.
*) Gelman, A. & Rubin, D. B. Inference from iterative simulation using multiple sequences. Stat. Sci. 7, 457–472 (1992).
Without necessarily getting into the code of it, but focusing more on the principles, I have a question about what I assume would be underfitting.
If I am training a network that recognizes true or false as to whether an image is of a dog, and I have maybe 40,000 images, where all dog images are labeled as 1, and all other images are labeled as 0 - what can I do to assure accuracy so that, if only maybe 5,000 of those images are dogs, the network does not act “lazily” from its training, and also label dogs as closer to 0 than 1?
For example, the main purpose of this question is to be able to recognize with high accuracy if an image really is of a dog, without really caring too much about the other images, other than the fact that they are not of dogs. Also, I would like to be able to retain the probability that the guess is correct, because this is highly important for my purposes.
The only two things I was able to come up with were to:
Have more nodes in the network, or
Have half of the images be of dogs (so use 10,000 images where 5,000 of them are dogs).
But I think this 2nd option might give dogs a disproportionately large chance of being the output of the testing data, which would destroy the accuracy and the whole purpose of this network.
I am sure this has been addressed before, so even a point in the right direction would be highly appreciated!
So you have a binary classification task where both classes appear with different frequency in your dataset. About 1/8 is "dog" and 7/8 is "no dog".
In order to avoid biased learning towards one or the other class, it is important that you stratify your training, validation and test data so that these fractions are kept across every subset.
You say that you want to "retain the probability" that the guess is correct - I assume you mean you want to evaluate the "dogness"-probability as output variable. That's a simple softmax output layer with two outputs: 1st is "dog", 2nd "not dog". It's the typical way to address classification problems, regardless of the number of classes you need to distinguish.
I tried to go through this paper which describes the ECT algorithm but could not make much out of it.
I know it is different from one-against-al (oaa) and even performs better than oaa.I wanted a simple explanation about how ect works.
ECT and Filter trees are useful (only) if you have a very big number of output labels (classes), let's say N=1000. With OAA (one-against-all), it would mean to do N binary classification tasks for each example (during both training and testing). With ECT you can make the prediction much faster: log(N). You can imagine Filter trees (which are the basis of ECT) as a decision tree where in each node you ask whether the example belongs to one set of labels or another set of labels (using all the features, unlike original decision trees).
In general, ECT is worse (in terms of loss or accuracy) than OAA (but in some cases it may be almost as good as OAA). With N=10 labels, you should try OAA first. With N>1000, OAA is too slow (and even the accuracy is low), you should try ECT (or --log_multi or --csoaa_ldf in VW, if you can preselect a smaller number of labels which are relevant for each example).
See http://cilvr.cs.nyu.edu/diglib/lsml/logarithmic.pdf
I have a picture.1200*1175 pixel.I want to train a net(mlp or hopfield) to learn a specific part of it(201*111pixel) to save its weight to use in a new net(with the same previous feature)only without train it to find that specific part.now there are this questions :what kind of nets is useful;mlp or hopfield,if mlp;the number of hidden layers;the trainlm function is unuseful because "out of memory" error.I convert the picture to a binary image,is it useful?
What exactly do you need the solution to do? Find an object with an image (like "Where's Waldo"?). Will the target object always be the same size and orientation? Might it look different because of lighting changes?
If you just need to find a fixed pattern of pixels within a larger image, I suggest using a straightforward correlation measure, such as crosscorrelation to find it efficiently.
If you need to contend with any of the issues mentioned above, then there are two basic solutions: 1. Build a model using examples of the object in different poses, scalings, etc. so that the model will recognize any of them, or 2. Develop a way to normalize the patch of pixels being examined, to minimize the effect of those distortions (like Hu's invariant moments). If nothing else, yuo'll want to perform some sort of data reduction to get the number of inputs down. Technically, you could also try a model which is invariant to rotations, etc., but I don't know how well those work. I suspect that they are more tempermental than traditional approaches.
I found AdaBoost to be helpful in picking out only important bits of an image. That, and resizing the image to something very tiny (like 40x30) using a Gaussian filter will speed it up and put weight on more of an area of the photo rather than on a tiny insignificant pixel.