I'm training a neural net using Keras in Python for time-series climate data (predicting value X at time t=T), and tried adding a (20%) dropout layer on the inputs, which seemed to limit overfitting and cause a slight increase in performance. However, after I added a new and particularly useful feature (the value of the response variable at time of prediction t=0), I found massively increased performance by removing the dropout layer. This makes sense to me, since I can imagine how the neural net would "learn" the importance of that one feature and base the rest of its training around adjusting that value (i.e, "how do these other features affect how the response at t=0 changes by time t=T").
In addition, there are a few other features that I think should be present for all epochs. That said, I am still hopeful that a dropout layer could improve the model performance-- it just needs to not drop out certain features, like X at t_0: I need a dropout layer that will only drop out certain features.
I have searched for examples of doing this, and read the Keras documentation here, but can't seem to find a way to do it. I may be missing something obvious, as I'm still not familiar with how to manually edit layers. Any help would be appreciated. Thanks!
Edit: sorry for any lack of clarity. Here is the code where I define the model (p is the number of features):
def create_model(p):
model = Sequential()
model.add(Dropout(0.2, input_shape=(p,))) # % of features dropped
model.add(Dense(1000, input_dim=p, kernel_initializer='normal'
, activation='sigmoid'))
model.add(Dense(30, kernel_initializer='normal', activation='relu'))
model.add(Dense(1, kernel_initializer='normal',activation='linear'))
model.compile(loss=cost_fn, optimizer='adam')
return model
The best way I can think of applying dropout only to specific features is to simply separate the features in different layers.
For that, I suggest you simply divide your inputs in essential features and droppable features:
from keras.layers import *
from keras.models import Model
def create_model(essentialP,droppableP):
essentialInput = Input((essentialP,))
droppableInput = Input((droppableP,))
dropped = Dropout(0.2)(droppableInput) # % of features dropped
completeInput = Concatenate()([essentialInput, dropped])
output = Dense(1000, kernel_initializer='normal', activation='sigmoid')(completeInput)
output = Dense(30, kernel_initializer='normal', activation='relu')(output)
output = Dense(1, kernel_initializer='normal',activation='linear')(output)
model = Model([essentialInput,droppableInput],output)
model.compile(loss=cost_fn, optimizer='adam')
return model
Train the model using two inputs. You have to manage your inputs before training:
model.fit([essential_train_data,droppable_train_data], predictions, ...)
I don't see any harm to using dropout in the input layer. The usage/effect would be a little different than normal of course. The effect would be similar to adding synthetic noise to an input signal; only the feature/pixel/whatever would be entirely unknown[zeroed out] instead of noisy. And inserting synthetic noise into the input is one of the oldest ways to improve robustness; certainly not bad practice as long as you think about whether it makes sense for your data set.
This question has already an accepted answer but it seems to me you are using dropout in a bad way.
Dropout is only for the hidden layers, not for the input layer !
Dropout act as a regularizer, and prevent the hidden layer complex coadaptation, quoting Hinton paper "Our work extends this idea by showing that dropout can be effectively applied in the hidden layers as well and that it can be interpreted as a form of model averaging" (http://www.jmlr.org/papers/volume15/srivastava14a/srivastava14a.pdf)
Dropout can be seen as training several different models with your data and averaging the prediction at test time. If you prevent your models to have all the inputs during training, it will perform badly, especially if one input is crucial. What you want is actually avoid overfitting, meaning you prevent too complex models during the training phase (so each of your models will select the most important features first) before testing.
It is common practice to drop some of the features in ensemble learning but it is control and not stochastic like for dropout. It also works for neural networks as hidden layers have (often) way more neurons as inputs, and so dropout follows the law of big numbers, as for a small number of inputs, you can have in some bad case almost all your inputs dropped.
In conlusion: it is a bad practice to use dropout in the input layer of a neural network.
Related
I have a binary classification problem. I use the following keras model to do my classification.
input1 = Input(shape=(25,6))
x1 = LSTM(200)(input1)
input2 = Input(shape=(24,6))
x2 = LSTM(200)(input2)
input3 = Input(shape=(21,6))
x3 = LSTM(200)(input3)
input4 = Input(shape=(20,6))
x4 = LSTM(200)(input4)
x = concatenate([x1,x2,x3,x4])
x = Dropout(0.2)(x)
x = Dense(200)(x)
x = Dropout(0.2)(x)
output = Dense(1, activation='sigmoid')(x)
However, the results I get is extremely bad. I thought the reason is that I have too many features, thus, needs have more improved layers after the concatenate.
I was also thinking if it would be helpful to used a flatten() layer after the concatenate.
anyway, since I am new to deep learning, I am not so sure how to make this a better model.
I am happy to provide more details if needed.
Here is what I can suggest
Remove every things that prevent overfitting, such as Dropout and regularizer. What can happen is that your model may not be able to capture the complexity of your data using given layer, so you need to make sure that your model is able to overfit first before adding regularizer.
Now try increase number of Dense layer and number of neuron in each layer until you can see some improvement. There is also a possibility that your data is too noisy or you have only few data to train the model so you can't even produce a useful predictions.
Now if you are LUCKY and you can see overfitting, you can add Dropout and regularizer.
Because every neural network is a gradient base algorithm, you may end up at local minimum. You may also need to run the algorithm multiple times with different initial weight before you can get a good result or You can change your loss function so that you have a convex problem where local minimum is global minimum.
If you can't achieve better result
You may need to try different topology because LSTM is just trying to model a system that assume to have Markov property. you can look at nested-LSTM or something like that, which model the system in the way that next time step is not just depend on current time step.
The Dropout right before the output layer could be problematic. I would suggest removing both Dropout layers and evaluating performance, then re-introduce regularization once the model is performing well on the the training set.
I'm training a LSTM Neural Network to predict a volatilty (timeseries) in Keras. At the moment, my network is specified as follows:
model = Sequential()
model.add(LSTM(10, input_shape=(1,1), kernel_regularizer = l2(0.0001)))
model.add(Dense(1, activation = 'relu'))
model.compile(loss='mean_squared_error', optimizer='adam')
model.fit(X_train, y_train, validation_split=0.2, epochs=100, batch_size=16)
Here, I have a lot of parameters I could cross validate:
units in LSTM?
More layers?
regularizer (L1 or l2, and amount)?
Activation function?
Optimizer?
Batch size?
However, CV on all these parameters would result in huge computational time, so how do I determind the correct specifications for all of them?
So far as I know, doing grid-search might be the best approach. However, you can lessen your search space by examining your data. If you don't have much data, try to go for a smaller model, don't go too big (or else it will overfit). This can lessen your search space a bit. Some say less layer but more unit works well for low-resource data, but still, it is not guaranteed.
Regularizer can sometimes good or bad, it depends on the task. You'll never know if the setting is correct or not unless you experiment on it.
For batch size, it is recommended to experiment on the batch size from 16 to 512 (or you can go higher if you can). The larger the batch size is, the faster it trains, the more memory it consumes. Smaller batch size also means the model will "walk" more random. In other words, the loss will decrease at a more random pace.
For optimizer, if you want to grid search, just use Adam. It is quite good for most tasks.
All in all, no one can guarantee that tuning different hyperparameters will result in a performance gain. It all needs to be experimented and record. That's why there are so many research papers done on hyperparameters tuning.
I have a dataset and i want to decide on which ML algorithm to apply to my given problem.
Customers are to fill out an assessment questionnaire of 50 questions. Examples of the questions are, what is your job, previous job history, how much do you earn, have you been rejected for a loan etc, and the end goal is to decide whether they should be rejected or not.
I have circa 500 entries for my algorithm to learn from and have pre-processed my dataset and converted the inputs into a numpy array and wondering what would be the best algorithm to use? Should i use a classification algorithm or a neural network in tensorflow and if the latter, what would be the layers I should use?
Thanks
How about beginning with xgboost or random forest? - So plain "old" ML?
The advantage would be that you could visualize the decision tree of the model once trained.
If using a NN in tensorflow (or even easier: keras with tensorflow backend), you could go with a MLP (multi layer perceptron), since the questions answers have fixed position in the input. You don't need many layers.
Important is that you normalize your input data columnwise, so that the input numbers are not much bigger/smaller than +1/-1, respectively. Introductory books often miss this point, though important.
Since your target labeling is "accept" or "reject", binary classifier will do it (also in the machine learning approach). (You use 0 and 1 as labels).
For NN, you don't need for such kind of classification that many layers or neurons. Try the smallest network first. let's say 10 neurons in first layer, then 7 neurons in the next layer (probably even less) and then 1 output neuron for the binary decision.
With Keras this would be:
from keras.models import Sequential
from keras.layers import Dense
def create_mlp(n_input = 500): # number of columns of input data 500 here
model = Sequential()
model.add(Dense(10, input_dim=n_input, kernel_initializer='normal', activation='relu')) # init = kernel_initializer
model.add(Dense(7, kernel_initializer='normal', activation='relu'))
model.add(Dense(1, activation='sigmoid'))
model.compile(optimizer='adam',
loss='binary_crossentropy',
metrics=['acc'])
return model
model = create_mlp(500) # this will generate the correct NN compiled.
Your data frame (or Numpy input array must have as rows the samples,
the columns are for each answer for a question 1 column.
the answers you have to encode in a numeric form. Numbers should be small - the best between -1 and 1. NNs don't like big numbers. Thus column-wise normalization can help.)
That's it. I learned all this stuff last year. Good luck for learning. It will be tons of fun!
I am attempting to use keras to build an activity classifier from accelerometer signals. However, I am experiencing extreme overfitting of the data even with the most simplistic of models.
The input data is of shape (10,3) and contains roughly .1 second of data from the accelerometer in 3 dimensions. The model is simply
model = Sequential()
model.add(Flatten(input_shape=(10,3)))
model.add(Dense(2, activation='softmax'))
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
The model should output the label [1,0] for walking activities and [0,1] for non-walking activities. After training I get 99.8% accuracy (if only it was real...). When I attempt to predict on data that wasn't used for training, I get 50% accuracy, verifying that the net isn't really "learning" anything except to predict a single class value.
The data is being prepared from 100hz triaxial accelerometer signals. I am not preprocessing the data in any way except for windowing it into bins on length 10 that overlap with the previous bin by 50%. What measures can I take to make the network produce actual predictions? I have tried increasing the window size but the results remain the same. Any advice/general tips are greatly appreciated.
Ian
Try adding some hidden layers and dropout layers to your network. You could create a simple Multi Layer Perceptron (MLP) with a couple of extra lines in between your Flatten layer and Dense layer:
model.add(Dense(64, activation='relu', input_dim=30))
model.add(Dropout(0.25))
model.add(Dense(64, activation='relu'))
model.add(Dropout(0.1))
Or check out this guide. which explains how to create a simple MLP.
Without any hidden layers your model will not actually be 'learning' from the input data, rather it will be mapping the input number of features to the output number of features.
The more layers you add, the more intermediate features and patterns it should extract from the input data which should lead to better model predictions for test data. There will be a lot of trial and error to design the best model as too many layers can result in over fitting.
You have not provided information about how you train the model so that may be the cause of the issue as well. You must ensure that the data is spit into training, testing and validation sets. Some possible split ratios to use for training, validation, test data are: 60%:20%:20%, or 70%:15%:15%. This is ultimately something that you must also decide.
The problem of overfitting was caused by the input data type. The values passed to the classifier should have been float values with 2 decimal places. Somewhere along the way, some of these values had been augmented and had significantly more than 2 decimal places. That is, the input should have looked like
[9.81, 10.22, 11.3]
but instead looked like
[9.81000000012, 10.220010431, 11.3000000101]
The classifier was making its prediction based on this feature, which is obviously not the desired behavior! Lessoned learned - make sure the data preparation is consistent! Thanks to #umutto for the suggestions of random forests, the simple structure was helpful for diagnosing purposes.
I'm trying to model a technical process (a number of nonlinear equations) with artificial neural networks. The function has a number of inputs and a number of outputs (e.g. 50 inputs, 150 outputs - all floats).
I have tried the python library ffnet (wrapper for a fortran library) with great success. The errors for a certain dataset are well below 0.2%.
It is using a fully connected graph and these additional parameters.
Basic assumptions and limitations:
Network has feed-forward architecture.
Input units have identity activation function, all other units have sigmoid activation function.
Provided data are automatically normalized, both input and output, with a linear mapping to the range (0.15, 0.85). Each input and output is treated separately (i.e. linear map is unique for each input and output).
Function minimized during training is a sum of squared errors of each output for each training pattern.
I am using one input layer, one hidden layer (size: 2/3 of input vector + size of output vector) and an output layer. I'm using the scipy conjugate gradient optimizer.
The downside of ffnet is the long training time and the lack of functionality to use GPUs. Therefore i want to switch to a different framework and have chosen keras with TensorFlow as the backend.
I have tried to model the previous configuration:
model = Sequential()
model.add(Dense(n_hidden, input_dim=n_in))
model.add(BatchNormalization())
model.add(Dense(n_hidden))
model.add(Activation('sigmoid'))
model.add(Dense(n_out))
model.add(Activation('sigmoid'))
model.summary()
model.compile(loss='mean_squared_error',
optimizer='Adamax',
metrics=['accuracy'])
However the results are far worse, the error is up to 0.5% with a few thousand (!) epochs of training. The ffnet training was automatically canceled at 292 epochs. Furthermore the differences between the network response and the validation target are not centered around 0, but mostly negative.
I have tried all optimizers and different loss functions. I have also skipped the BatchNormalization and normalized the data manually in the same way that ffnet does it. Nothing helps.
Does anyone have a suggestion to obtain better results with keras?
I understand you are trying to re-train the same architecture from scratch, with a different library. The first fundamental issue to keep in mind here is that neural nets are not necessarily reproducible, when weights are initialized randomly.
For example, here is the default constructor parameter for Dense in Keras:
init='glorot_uniform'
But even before trying to evaluate the convergence of Keras optimizations, I would recommend trying to port the weights for which you got good results, from ffnet, into your Keras model. You can do so either with the kwarg Dense(..., weights=) of each layer, or globally at the end model.set_weights(...)
Using the same weights must yield the exact same result between the two libs. Unless you run into some floating point rounding issues. I believe that as long as porting the weights is not consistent, working on the optimization is unlikely to help.