I have tried to train simple backpropagation neural network with the xor function. When I use tanh(x) as activation function, with the derivative 1-tanh(x)^2, I get the right result after about 1000 iterations. However, when I use g(x) = 1/(1+e^(-x)) as an activation function, with the derivative g(x)*(1-g(x)), I need about 50000 iterations to get the right result. What can be the reason?
Thank you.
Yes, what you observe is true. I have similar observations when training neural networks using back propagations. For XOR problem, I used to set up a 2x20x2 network, logistic function takes 3000+ episodes to get below result:
[0, 0] -> [0.049170633762142486]
[0, 1] -> [0.947292007836417]
[1, 0] -> [0.9451808598939389]
[1, 1] -> [0.060643862846171494]
While using tanh as activation function, here is the result after 800 episodes. tanh converges consistently faster than logistic.
[0, 0] -> [-0.0862215901296476]
[0, 1] -> [0.9777578145233919]
[1, 0] -> [0.9777632805205176]
[1, 1] -> [0.12637838259658932]
The two functions' shape look like below (credit: efficient backprop):
The left is the standard logistic function: 1/(1+e^(-x)).
The right is the tanh function, also known as hyperbolic tangent.
It's easy to see that tanh is antisymmetric about the origin.
According to efficient Backprop,
Symmetric sigmoids such as tanh often converge faster than standard logistic function.
Also from wiki Logistic regression:
Practitioners caution that sigmoidal functions which are antisymmetric about the origin (e.g. the hyperbolic tangent) lead to faster convergence when training networks with backpropagation.
See efficient Backprop for more details explaining the intuition here.
See elliott for an alternative of tanh with easier computations. It's shown below as the black curve (the blue one is the original tanh).
Two things should stand out from the above chart. First, TANH usually needed fewer iterations to train than Elliott. So the training accuracy is not as good with Elliott, for an Encoder. However, notice the training times. Elliott completed its entire task, even with the extra iterations it had to do, in half the time of TANH. This is a huge improvement and literally means that in this case, Elliott will cut your training time in half, and deliver the same final training error. While it does take more training iterations to get there, the speed per iteration is so much faster it still results in the training time being cut in half.
Related
I am training a unsupervised NN model and for some reason, after exactly one epoch (80 steps), model stops learning.
]
Do you have any idea why it might happen and what should I do to prevent it?
This is more info about my NN:
I have a deep NN that tries to solve an optimization problem. My loss function is customized and it is my objective function in the optimization problem.
So if my optimization problems is min f(x) ==> loss, now in my DNN loss = f(x). I have 64 input, 64 output, 3 layers in between :
self.l1 = nn.Linear(input_size, hidden_size)
self.relu1 = nn.LeakyReLU()
self.BN1 = nn.BatchNorm1d(hidden_size)
and last layer is:
self.l5 = nn.Linear(hidden_size, output_size)
self.tan5 = nn.Tanh()
self.BN5 = nn.BatchNorm1d(output_size)
to scale my network.
with more layers and nodes(doubles: 8 layers each 200 nodes), I can get a little more progress toward lower error, but again after 100 steps training error becomes flat!
The symptom is that the training loss stops being improved relatively early. Suppose that your problem is learnable at all, there are many reasons for the for this behavior. Following are most relavant:
Improper preprocessing of input: Neural network prefers input with
zero mean. E.g., if the input is all positive, it will restrict the
weights to be updated in the same direction, which may not be
desirable (https://youtu.be/gYpoJMlgyXA).
Therefore, you may want to subtract the mean from all the images (e.g., subtract 127.5 from each of the 3 channels). Scaling to make unit standard deviation in each channel may also be helpful.
Generalization ability of the network: The network is not complicated
or deep enough for the task.
This is very easy to check. You can train the network on just a few
images (says from 3 to 10). The network should be able to overfit the
data and drives the loss to almost 0. If it is not the case, you may
have to add more layers such as using more than 1 Dense layer.
Another good idea is to used pre-trained weights (in applications of Keras documentation). You may adjust the Dense layers at the top to fit with your problem.
Improper weight initialization. Improper weight initialization can
prevent the network from converging (https://youtu.be/gYpoJMlgyXA,
the same video as before).
For the ReLU activation, you may want to use He initialization
instead of the default Glorot initialiation. I find that this may be
necessary sometimes but not always.
Lastly, you can use debugging tools for Keras such as keras-vis, keplr-io, deep-viz-keras. They are very useful to open the blackbox of convolutional networks.
I faced the same problem then I followed the following:
After going through a blog post, I managed to determine that my problem resulted from the encoding of my labels. Originally I had them as one-hot encodings which looked like [[0, 1], [1, 0], [1, 0]] and in the blog post they were in the format [0 1 0 0 1]. Changing my labels to this and using binary crossentropy has gotten my model to work properly. Thanks to Ngoc Anh Huynh and rafaelvalle!
I am developing a model using linear regression to predict the age. I know that the age is from 0 to 100 and it is a possible value. I used conv 1 x 1 in the last layer to predict the real value. Do I need to add a ReLU function after the output of convolution 1x1 to guarantee the predicted value is a positive value? Currently, I did not add ReLU and some predicted value becomes negative value like -0.02 -0.4…
There's no compelling reason to use an activation function for the output layer; typically you just want to use a reasonable/suitable loss function directly with the penultimate layer's output. Specifically, a RELU doesn't solve your problem (or at most only solves 'half' of it) since it can still predict above 100. In this case -predicting a continuous outcome- there's a few standard loss functions like squared error or L1-norm.
If you really want to use an activation function for this final layer and are concerned about always predicting within a bounded interval, you could always try scaling up the sigmoid function (to between 0 and 100). However, there's nothing special about sigmoid here - any bounded function, ex. any CDF of a signed, continuous random variable, could be similarly used. Though for optimization, something easily differentiable is important.
Why not start with something simple like squared-error loss? It's always possible to just 'clamp' out-of-range predictions to within [0-100] (we can give this a fancy name like 'doubly RELU') when you need to actually make predictions (as opposed to during training/testing), but if you're getting lots of such errors, the model might have more fundamental problems.
Even for a regression problem, it can be good (for optimisation) to use a sigmoid layer before the output (giving a prediction in the [0:1] range) followed by a denormalization (here if you think maximum age is 100, just multiply by 100)
This tip is explained in this fast.ai course.
I personally think these lessons are excellent.
You should use a sigmoid activation function, and then normalize the targets outputs to the [0, 1] range. This solves both issues of being positive and with a limit.
You can easily then denormalize the neural network outputs to get an output in the [0, 100] range.
I implemented a binary Logistic Regression classifier. Just to play, around I replaced the sigmoid function (1 / 1 + exp(-z)), with tanh. The results were exactly the same, with the same 0.5 threshold for classification and even though tanh is in the range {-1,1} while sigmoid is in the range {0,1}.
Does it really matter that we use the sigmoid function or can any differentiable non-linear function like tanh work?
Thanks.
Did you also change the function in the training, or you just used the same training method and then changed the sigmoid to tanh?
I think what has very likely happened is the following. Have a look at the graphs of sigmoid and tanh:
sigmoid: http://www.wolframalpha.com/input/?i=plot+sigmoid%28x%29+for+x%3D%28-1%2C+1%29
tanh: http://www.wolframalpha.com/input/?i=plot+tanh%28x%29+for+x%3D%28-1%2C+1%29
We can see that in the tanh case, the value y = 0.5 is around x = 0.5. In the sigmoid, the x = 0.5 gets us roughly y = 0.62. Therefore, what I think has probably happened now is that your data doesn't contain any point that would fall within this range, hence you get exactly the same results. Try printing the sigmoid values for your data and see if there is any between 0.5 and 0.62.
The reason behind using the sigmoid function is that it is derived from probability and maximum likelihood. While the other functions may work very similarly, they will lack this probabilistic theory background. For details see for example http://luna.cas.usf.edu/~mbrannic/files/regression/Logistic.html or http://www.cs.cmu.edu/~tom/mlbook/NBayesLogReg.pdf
The range of the function should be {0,1} as it represents probability of the outcome.
I am moving my first steps in neural networks and to do so I am experimenting with a very simple single layer, single output perceptron which uses a sigmoidal activation function. I am updating my weights on-line each time a training example is presented using:
weights += learningRate * (correct - result) * {input,1}
Here weights is a n-length vector which also contains the weight from the bias neuron (- threshold), result is the result as computed by the perceptron (and processed using the sigmoid) when given the input, correct is the correct result and {input,1} is the input augmented with 1 (the fixed input from the bias neuron). Now, when I try to train the perceptron to perform logic AND, the weights don't converge for a long time, instead they keep growing similarly and they maintain a ratio of circa -1.5 with the threshold, for instance the three weights are in sequence:
5.067160008240718 5.105631826680446 -7.945513136885797
...
8.40390853077094 8.43890306970281 -12.889540730182592
I would expect the perceptron to stop at 1, 1, -1.5.
Apart from this problem, which looks like connected to some missing stopping condition in the learning, if I try to use the identity function as activation function, I get weight values oscillating around:
0.43601272528257057 0.49092558197172703 -0.23106430854347537
and I obtain similar results with tanh. I can't give an explanation to this.
Thank you
Tunnuz
It is because the sigmoid activation function doesn't reach one (or zero) even with very highly positive (or negative) inputs. So (correct - result) will always be non-zero, and your weights will always get updated. Try it with the step function as the activation function (i.e. f(x) = 1 for x > 0, f(x) = 0 otherwise).
Your average weight values don't seem right for the identity activation function. It might be that your learning rate is a little high -- try reducing it and see if that reduces the size of the oscillations.
Also, when doing online learning (aka stochastic gradient descent), it is common practice to reduce the learning rate over time so that you converge to a solution. Otherwise your weights will continue to oscillate.
When trying to analyze the behavior of the perception, it helps to also look at correct and result.
Is it a good practice to use sigmoid or tanh output layers in Neural networks directly to estimate probabilities?
i.e the probability of given input to occur is the output of sigmoid function in the NN
EDIT
I wanted to use neural network to learn and predict the probability of a given input to occur..
You may consider the input as State1-Action-State2 tuple.
Hence the output of NN is the probability that State2 happens when applying Action on State1..
I Hope that does clear things..
EDIT
When training NN, I do random Action on State1 and observe resultant State2; then teach NN that input State1-Action-State2 should result in output 1.0
First, just a couple of small points on the conventional MLP lexicon (might help for internet searches, etc.): 'sigmoid' and 'tanh' are not 'output layers' but functions, usually referred to as "activation functions". The return value of the activation function is indeed the output from each layer, but they are not the output layer themselves (nor do they calculate probabilities).
Additionally, your question recites a choice between two "alternatives" ("sigmoid and tanh"), but they are not actually alternatives, rather the term 'sigmoidal function' is a generic/informal term for a class of functions, which includes the hyperbolic tangent ('tanh') that you refer to.
The term 'sigmoidal' is probably due to the characteristic shape of the function--the return (y) values are constrained between two asymptotic values regardless of the x value. The function output is usually normalized so that these two values are -1 and 1 (or 0 and 1). (This output behavior, by the way, is obviously inspired by the biological neuron which either fires (+1) or it doesn't (-1)). A look at the key properties of sigmoidal functions and you can see why they are ideally suited as activation functions in feed-forward, backpropagating neural networks: (i) real-valued and differentiable, (ii) having exactly one inflection point, and (iii) having a pair of horizontal asymptotes.
In turn, the sigmoidal function is one category of functions used as the activation function (aka "squashing function") in FF neural networks solved using backprop. During training or prediction, the weighted sum of the inputs (for a given layer, one layer at a time) is passed in as an argument to the activation function which returns the output for that layer. Another group of functions apparently used as the activation function is piecewise linear function. The step function is the binary variant of a PLF:
def step_fn(x) :
if x <= 0 :
y = 0
if x > 0 :
y = 1
(On practical grounds, I doubt the step function is a plausible choice for the activation function, but perhaps it helps understand the purpose of the activation function in NN operation.)
I suppose there an unlimited number of possible activation functions, but in practice, you only see a handful; in fact just two account for the overwhelming majority of cases (both are sigmoidal). Here they are (in python) so you can experiment for yourself, given that the primary selection criterion is a practical one:
# logistic function
def sigmoid2(x) :
return 1 / (1 + e**(-x))
# hyperbolic tangent
def sigmoid1(x) :
return math.tanh(x)
what are the factors to consider in selecting an activation function?
First the function has to give the desired behavior (arising from or as evidenced by sigmoidal shape). Second, the function must be differentiable. This is a requirement for backpropagation, which is the optimization technique used during training to 'fill in' the values of the hidden layers.
For instance, the derivative of the hyperbolic tangent is (in terms of the output, which is how it is usually written) :
def dsigmoid(y) :
return 1.0 - y**2
Beyond those two requriements, what makes one function between than another is how efficiently it trains the network--i.e., which one causes convergence (reaching the local minimum error) in the fewest epochs?
#-------- Edit (see OP's comment below) ---------#
I am not quite sure i understood--sometimes it's difficult to communicate details of a NN, without the code, so i should probably just say that it's fine subject to this proviso: What you want the NN to predict must be the same as the dependent variable used during training. So for instance, if you train your NN using two states (e.g., 0, 1) as the single dependent variable (which is obviously missing from your testing/production data) then that's what your NN will return when run in "prediction mode" (post training, or with a competent weight matrix).
You should choose the right loss function to minimize.
The squared error does not lead to the maximum likelihood hypothesis here.
The squared error is derived from a model with Gaussian noise:
P(y|x,h) = k1 * e**-(k2 * (y - h(x))**2)
You estimate the probabilities directly. Your model is:
P(Y=1|x,h) = h(x)
P(Y=0|x,h) = 1 - h(x)
P(Y=1|x,h) is the probability that event Y=1 will happen after seeing x.
The maximum likelihood hypothesis for your model is:
h_max_likelihood = argmax_h product(
h(x)**y * (1-h(x))**(1-y) for x, y in examples)
This leads to the "cross entropy" loss function.
See chapter 6 in Mitchell's Machine Learning
for the loss function and its derivation.
There is one problem with this approach: if you have vectors from R^n and your network maps those vectors into the interval [0, 1], it will not be guaranteed that the network represents a valid probability density function, since the integral of the network is not guaranteed to equal 1.
E.g., a neural network could map any input form R^n to 1.0. But that is clearly not possible.
So the answer to your question is: no, you can't.
However, you can just say that your network never sees "unrealistic" code samples and thus ignore this fact. For a discussion of this (and also some more cool information on how to model PDFs with neural networks) see contrastive backprop.