Giving less weight to data coming from another dataset that is noisy - tensorflow

I have two datasets, one with clean data and one with dirty data. I train a Roberta model on the clean dataset and then get predictions for the dirty dataset. Those predictions with a probability greater than 0.9 go to the clean dataset. I then retrain the Roberta model with this new dataset (clean + dirty moving to clean).
For the retraining I am using the MAE loss function (more robust to noisy labels) and I use weights to give less value to the data that passes from the dirty to the clean dataset, as follows:
loss = torch.mean(torch.abs(y_true - y_pred) * weights)
Initially I am using an arbitrary weight of 0.5 for all the dirty data that gets passed into the clean dataset. However, I would like to assign them a weight in a more academic way, not so arbitrary.
How can I do that?

One way to choose the weight is based on your confidence in the dirty data and assign the weight accordingly. For example, if you think that 90% of dirty data is labeled correctly, then choosing 0.9 as the weight for the noisy data is a reasonable option.
Additionally, there is a whole literature on learning from noisy labels, you can check this survey for more information: https://arxiv.org/abs/2007.08199

Out of curiosity, why not use cleanlab to find the label errors and other data issues in your dataset for you directly? https://github.com/cleanlab/cleanlab
It handles most data issues for ML in a few lines of code, some examples:
Find label issues in 1 line of code
from cleanlab.classification import CleanLearning
from cleanlab.filter import find_label_issues
# Option 1 - works with sklearn-compatible models - just input the data and labels ツ
label_issues_info = CleanLearning(clf=sklearn_compatible_model).find_label_issues(data, labels)
# Option 2 - works with ANY ML model - just input the model's predicted probabilities
ordered_label_issues = find_label_issues(
labels=labels,
pred_probs=pred_probs, # out-of-sample predicted probabilities from any model
return_indices_ranked_by='self_confidence',
)
Train a model as if the dataset did not have errors -- 3 lines of code
from sklearn.linear_model import LogisticRegression
from cleanlab.classification import CleanLearning
cl = CleanLearning(clf=LogisticRegression()) # any sklearn-compatible classifier
cl.fit(train_data, labels)
# Estimate the predictions you would have gotten if you trained without mislabeled data.
predictions = cl.predict(test_data)
Journal of AI Research (with theory to prove it works): https://arxiv.org/abs/1911.00068publication
errors found using cleanlab: https://labelerrors.com/
Documentation and runnable tutorials for cleanlab: https://docs.cleanlab.ai/

Related

Keras LSTM: how to predict beyond validation vs predictions?

When dealing with time series forecasting, I've seen most people follow these steps when using an LSTM model:
Obtain, clean, and pre-process data
Take out validation dataset for future comparison with model predictions
Initialise and train LSTM model
Use a copy of validation dataset to be pre-processed exactly like the training data
Use trained model to make predictions on the transformed validation data
Evaluate results: predictions vs validation
However, if the model is accurate, how do you make predictions that go beyond the end of the validation period?
The following only accepts data that have been transformed in the same way as the training data, but for predictions that go beyond the validation period, you don't have any input data to feed to the model. So, how do people do this?
# Predictions vs validation
predictions = model.predict(transformed_validation)
# Future predictions
future_predictions = model.predict(?)
To predict the ith value, your LSTM model need last N values.
So if you want to forecast, you should use each prediction to predict the next one.
In other terms you have to loop over something like
prediction = model.predict(X[-N:])
X.append(prediction)
As you can guess, you add your output in your input that's why your predictions can diverge and amplify uncertainty.
Other model are more stable to predict far future.
You have to break your data into training and testing and then fit your mode. Finally, you make a prediction like this.
future_predictions = model.predict(X_test)
Check out the link below for all details.
Time-Series Forecasting: Predicting Stock Prices Using An LSTM Model

model.evaluate() varies wildly with number of steps when using generators

Running tensorflow 2.x in Colab with its internal keras version (tf.keras). My model is a 3D convolutional UNET for multiclass segmentation (not sure if it's relevant).
I've successfully trained (high enough accuracy on validation) this model the traditional way but I'd like to do augmentation to improve it, therefore I'm switching to (hand-written) generators. When I use generators I see my loss increasing and my accuracy decreasing a lot (e.g.: loss increasing 4-fold, not some %) in the fit.
To try to localize the issue I've tried loading my trained weights and computing the metrics on the data returned by the generators. And what's happening makes no sense. I can see that the results visually are ok.
model.evaluate(validationGenerator,steps=1)
2s 2s/step - loss: 0.4037 - categorical_accuracy: 0.8716
model.evaluate(validationGenerator,steps=2)
2s/step - loss: 1.7825 - categorical_accuracy: 0.7158
model.evaluate(validationGenerator,steps=4)
7s 2s/step - loss: 1.7478 - categorical_accuracy: 0.7038
Why would the loss vary with the number of steps? I could guess some % due to statistical variations... not 4 fold increase!
If I try
x,y = next(validationGenerator)
nSamples = x.shape[0]
meanLoss = np.zeros(nSamples)
meanAcc = np.zeros(nSamples)
for pIdx in range(nSamples):
y_pred = model.predict(np.expand_dims(x[pIdx,:,:,:,:],axis=0))
meanAcc[pIdx]=np.mean(tf.keras.metrics.categorical_accuracy(np.expand_dims(y[pIdx,:,:,:,:],axis=0),y_pred))
meanLoss[pIdx]=np.mean(tf.keras.metrics.categorical_crossentropy(np.expand_dims(y[pIdx,:,:,:,:],axis=0),y_pred))
print(np.mean(meanAcc))
print(np.mean(meanLoss))
I get accuracy~85% and loss ~0.44. Which is what I expect from the previous fit, and it varies by vary little from one batch to the other. And these are the same exact numbers that I get if I do model.evaluate() with 1 step (using the same generator function).
However I need about 30 steps to run trough my whole training dataset. What should I do?
If I fit my already good model to this generator it indeed worsen the performances a lot (it goes from a nice segmentation of the image to uniform predictions of 25% for each of the 4 classes!!!!)
Any idea on where to debud the issue? I've also visually looked at the images produced by the generator and at the model predictions and everything looks correct (as testified by the numbers I found when evaluating using a single step). I've tried writing a minimal working example with a 2 layers model but... in it the issue does not happen.
UPDATE: Generators code
So, as I've been asked, these are the generators code. They're handwritten
def dataGen (X,Y_train):
patchS = 64 #set the size of the patch I extract
batchS = 16 #number of samples per batch
nSamples = X.shape[0] #get total number of samples
immSize = X.shape[1:] #get the shape of the iamge to crop
#Get 4 patches from each image
#extract them randomly, and in random patient order
patList = np.array(range(0,nSamples),dtype='int16')
patList = patList.reshape(nSamples,1)
patList = np.tile(patList,(4,2))
patList[:nSamples,0]=0 #Use this index to tell the code where to get the patch from
patList[nSamples:2*nSamples,0]=1
patList[2*nSamples:3*nSamples,0]=2
patList[3*nSamples:4*nSamples,0]=3
np.random.shuffle(patList)
patStart=0
Xout = np.zeros((batchS,patchS,patchS,patchS,immSize[3])) #allocate output vector
while True:
Yout = np.zeros((batchS,patchS,patchS,patchS)) #allocate vector of labels
for patIdx in range(batchS):
XSR = 32* (patList[patStart+patIdx,0]//2) #get the index of where to extract the patch
YSR = 32* (patList[patStart+patIdx,0]%2)
xStart = random.randrange(XSR,XSR+32) #get a patch randomly somewhere between a range
yStart = random.randrange(YSR,YSR+32)
zStart = random.randrange(0,26)
patInd = patList[patStart+patIdx,1]
Xout[patIdx,:,:,:,:] = X[patInd,xStart:(xStart+patchS),yStart:(yStart+patchS),zStart:(zStart+patchS),:]
Yout[patIdx,:,:,:] = Y_train[patInd,xStart:(xStart+patchS),yStart:(yStart+patchS),zStart:(zStart+patchS)]
if((patStart+patIdx)>(patList.shape[0]-2)):
np.random.shuffle(patList) #after going through the whole list restart
patStart=0
patStart = patStart+batchS
Yout = tf.keras.utils.to_categorical (Yout, num_classes=4, dtype='float32') #convert to one hot encoding
yield Xout, Yout
Posting the workaround I've found for the future person coming here from google.
Apparently the issue lies in how keras calls a handwritten generator. When it was called multiple times in a row by using evaluate(gen, steps=N) apparently it returned wrong outputs. There's no documentation around about how to address this or how a generator should be written.
I ended up writing my code using a tf.keras.utils.sequence class and the same previous code now works perfectly. No way to know why.
Here are different factors that affect loss & accuracy:
For Accuracy, we know that it measures the accuracy of the prediction: i.e. correct-classes /total-classes.
While loss tracks the inverse-confidence of the prediction.
A high Loss indicates that although the model is performing well with the prediction, It is becoming uncertain of the prediction it is making.
For example, For an image classification scenario, The image of a cat is passed into two models. Model A predicts {cat: 0.8, dog: 0.2} and model B predicts {cat: 0.6, dog: 0.4}.
Both models will score the same accuracy, but model B will have a higher loss.
On your evaluation part, Based on the documentation
Steps: Integer or None. Total number of steps (batches of samples) before declaring the evaluation round finished. Ignored with the default value of None. If x is a tf.data dataset and steps is None, 'evaluate' will run until the dataset is exhausted. This argument is not supported by array inputs.
So for simplify, it's getting the Nth batch of your validation samples.
It could be that the model prediction is becoming uncertain since the majority of the unknown data falls on those specific steps. which in your case, steps 2 & 3.
So, As the evaluation steps progress, The prediction becomes more uncertain leading to a higher loss.
You might need to retrain your model with more training samples but of course, you need to be careful since you might encounter overfitting.
In terms of data augmentation, you might wanna check this link
In Training Perspective, proper data augmentation is one of the factors that leads to good model performance.

MLP output of first layer is zero after one epoch

I've been running into an issue lately trying to train a simple MLP.
I'm basically trying to get a network to map the XYZ position and RPY orientation of the end-effector of a robot arm (6-dimensional input) to the angle of every joint of the robot arm to reach that position (6-dimensional output), so this is a regression problem.
I've generated a dataset using the angles to compute the current position, and generated datasets with 5k, 500k and 500M sets of values.
My issue is the MLP I'm using doesn't learn anything at all. Using Tensorboard (I'm using Keras), I've realized that the output of my very first layer is always zero (see image 1), no matter what I try.
Basically, my input is a shape (6,) vector and the output is also a shape (6,) vector.
Here is what I've tried so far, without success:
I've tried MLPs with 2 layers of size 12, 24; 2 layers of size 48, 48; 4 layers of size 12, 24, 24, 48.
Adam, SGD, RMSprop optimizers
Learning rates ranging from 0.15 to 0.001, with and without decay
Both Mean Squared Error (MSE) and Mean Absolute Error (MAE) as the loss function
Normalizing the input data, and not normalizing it (the first 3 values are between -3 and +3, the last 3 are between -pi and pi)
Batch sizes of 1, 10, 32
Tested the MLP of all 3 datasets of 5k values, 500k values and 5M values.
Tested with number of epoches ranging from 10 to 1000
Tested multiple initializers for the bias and kernel.
Tested both the Sequential model and the Keras functional API (to make sure the issue wasn't how I called the model)
All 3 of sigmoid, relu and tanh activation functions for the hidden layers (the last layer is a linear activation because its a regression)
Additionally, I've tried the very same MLP architecture on the basic Boston housing price regression dataset by Keras, and the net was definitely learning something, which leads me to believe that there may be some kind of issue with my data. However, I'm at a complete loss as to what it may be as the system in its current state does not learn anything at all, the loss function just stalls starting on the 1st epoch.
Any help or lead would be appreciated, and I will gladly provide code or data if needed!
Thank you
EDIT:
Here's a link to 5k samples of the data I'm using. Columns B-G are the output (angles used to generate the position/orientation) and columns H-M are the input (XYZ position and RPY orientation). https://drive.google.com/file/d/18tQJBQg95ISpxF9T3v156JAWRBJYzeiG/view
Also, here's a snippet of the code I'm using:
df = pd.read_csv('kinova_jaco_data_5k.csv', names = ['state0',
'state1',
'state2',
'state3',
'state4',
'state5',
'pose0',
'pose1',
'pose2',
'pose3',
'pose4',
'pose5'])
states = np.asarray(
[df.state0.to_numpy(), df.state1.to_numpy(), df.state2.to_numpy(), df.state3.to_numpy(), df.state4.to_numpy(),
df.state5.to_numpy()]).transpose()
poses = np.asarray(
[df.pose0.to_numpy(), df.pose1.to_numpy(), df.pose2.to_numpy(), df.pose3.to_numpy(), df.pose4.to_numpy(),
df.pose5.to_numpy()]).transpose()
x_train_temp, x_test, y_train_temp, y_test = train_test_split(poses, states, test_size=0.2)
x_train, x_val, y_train, y_val = train_test_split(x_train_temp, y_train_temp, test_size=0.2)
mean = x_train.mean(axis=0)
x_train -= mean
std = x_train.std(axis=0)
x_train /= std
x_test -= mean
x_test /= std
x_val -= mean
x_val /= std
n_epochs = 100
n_hidden_layers=2
n_units=[48, 48]
inputs = Input(shape=(6,), dtype= 'float32', name = 'input')
x = Dense(units=n_units[0], activation=relu, name='dense1')(inputs)
for i in range(1, n_hidden_layers):
x = Dense(units=n_units[i], activation=activation, name='dense'+str(i+1))(x)
out = Dense(units=6, activation='linear', name='output_layer')(x)
model = Model(inputs=inputs, outputs=out)
optimizer = SGD(lr=0.1, momentum=0.4)
model.compile(optimizer=optimizer, loss='mse', metrics=['mse', 'mae'])
history = model.fit(x_train,
y_train,
epochs=n_epochs,
verbose=1,
validation_data=(x_test, y_test),
batch_size=32)
Edit 2
I've tested the architecture with a random dataset where the input was a (6,) vector where input[i] is a random number and the output was a (6,) vector with output[i] = input[i]² and the network didn't learn anything. I've also tested a random dataset where the input was a random number and the output was a linear function of the input, and the loss converged to 0 pretty quickly. In short, it seems the simple architecture is unable to map a non-linear function.
the output of my very first layer is always zero.
This typically means that the network does not "see" any pattern in the input at all, which causes it to always predict the mean of the target over the entire training set, regardless of input. Your output is in the range of -𝜋 to 𝜋 probably with an expected value of 0, so it checks out.
My guess is that the model is too small to represent the data efficiently. I would suggest that you increase the number of parameters in the model by a factor of 10 or 100 and see if it starts seeing something. Limiting the number of parameters has a regularizing effect on the network, and strong regularization usually leads the the aforementioned derping to the mean.
I'm by no means a robotics expert, but I guess that there are a lot of situations where a small nudge in the output parameters causes a large change of the input. Let's say I'm trying to scratch my back with my left hand - the farther my hand goes to the left, the harder the task becomes, so at some point I might want to switch hands, which is a discontinuous configuration change. A bad analogy, sure, but I hope it demonstrates my hunch that there are certain places in the configuration space where small target changes cause large configuration changes.
Such large changes will cause a very large, very noisy gradient around those points. I'm not sure how well the network will work around these noisy gradients, but I would suggest as an experiment that you try to limit the training dataset to a set of outputs that are connected smoothly to one another in the configuration space of the arm, if that makes sense. Going further, you should remove any points from the dataset that are close to such configuration boundaries. To make up for that at inference time, you might instead want to sample several close-by points and choose the most common prediction as the final result. Hopefully some of those points will land in a smooth configuration area.
Also, adding batch normalization before each dense layer will help smooth the gradient and provide for more reliable training.
As for the rest of your hyperparameters:
A batch size of 32 is good, a very small batch size will make the gradient too noisy
The loss function is not critical, both MSE and MAE should work
The activation functions aren't critical, ReLU is a good default choice.
The default initializers a good enough.
Normalizing is important for Dense layers, so keep it
Train for as many epochs as you need as long as both the training and validation loss are dropping. If the validation loss hasn't dropped for 5-10 epochs you might as well stop early.
Adam is a good default choice. Start with a small learning rate and increase the learning rate at the beginning of training only if the training loss is dropping consistently over several epochs.
Further reading: 37 Reasons why your Neural Network is not working
I ended up replacing the first dense layer with a Conv1D layer and the network now seems to be learning decently. It's overfitting to my data, but that's territory I'm okay with.
I'm closing the thread for now, I'll spend some time playing with the architecture.

DeepLearning Anomaly Detection for images

I am still relatively new to the world of Deep Learning. I wanted to create a Deep Learning model (preferably using Tensorflow/Keras) for image anomaly detection. By anomaly detection I mean, essentially a OneClassSVM.
I have already tried sklearn's OneClassSVM using HOG features from the image. I was wondering if there is some example of how I can do this in deep learning. I looked up but couldn't find one single code piece that handles this case.
The way of doing this in Keras is with the KerasRegressor wrapper module (they wrap sci-kit learn's regressor interface). Useful information can also be found in the source code of that module. Basically you first have to define your Network Model, for example:
def simple_model():
#Input layer
data_in = Input(shape=(13,))
#First layer, fully connected, ReLU activation
layer_1 = Dense(13,activation='relu',kernel_initializer='normal')(data_in)
#second layer...etc
layer_2 = Dense(6,activation='relu',kernel_initializer='normal')(layer_1)
#Output, single node without activation
data_out = Dense(1, kernel_initializer='normal')(layer_2)
#Save and Compile model
model = Model(inputs=data_in, outputs=data_out)
#you may choose any loss or optimizer function, be careful which you chose
model.compile(loss='mean_squared_error', optimizer='adam')
return model
Then, pass it to the KerasRegressor builder and fit with your data:
from keras.wrappers.scikit_learn import KerasRegressor
#chose your epochs and batches
regressor = KerasRegressor(build_fn=simple_model, nb_epoch=100, batch_size=64)
#fit with your data
regressor.fit(data, labels, epochs=100)
For which you can now do predictions or obtain its score:
p = regressor.predict(data_test) #obtain predicted value
score = regressor.score(data_test, labels_test) #obtain test score
In your case, as you need to detect anomalous images from the ones that are ok, one approach you can take is to train your regressor by passing anomalous images labeled 1 and images that are ok labeled 0.
This will make your model to return a value closer to 1 when the input is an anomalous image, enabling you to threshold the desired results. You can think of this output as its R^2 coefficient to the "Anomalous Model" you trained as 1 (perfect match).
Also, as you mentioned, Autoencoders are another way to do anomaly detection. For this I suggest you take a look at the Keras Blog post Building Autoencoders in Keras, where they explain in detail about the implementation of them with the Keras library.
It is worth noticing that Single-class classification is another way of saying Regression.
Classification tries to find a probability distribution among the N possible classes, and you usually pick the most probable class as the output (that is why most Classification Networks use Sigmoid activation on their output labels, as it has range [0, 1]). Its output is discrete/categorical.
Similarly, Regression tries to find the best model that represents your data, by minimizing the error or some other metric (like the well-known R^2 metric, or Coefficient of Determination). Its output is a real number/continuous (and the reason why most Regression Networks don't use activations on their outputs). I hope this helps, good luck with your coding.

tf.contrib.learn.LinearRegressor builds unexpectedly bad model for a data with one feature

I am building a simple linear regressor for the data from the csv. Data includes weight and height values of some people. Overall learning process is very simple:
MAX_STEPS = 2000
# ...
features = [tf.contrib.layers.real_valued_column(feature_name) for feature_name in FEATURES_COL]
# ...
linear_regressor = tf.contrib.learn.LinearRegressor(feature_columns=features)
linear_regressor.fit(input_fn=prepare_input, max_steps=MAX_STEPS)
However, the model that is built by the regressor is, unexpectedly, bad. Result could be illustrated with the next picture:
Visualization code(just in case):
plt.plot(height_and_weight_df_filtered[WEIGHT_COL],
linear_regressor.predict(input_fn=prepare_full_input),
color='blue',
linewidth=3)
Here is the same data been given to the LinearRegression class from the scikit-learn:
lr_updated = linear_model.LinearRegression()
lr_updated.fit(weight_filtered_reshaped, height_filtered)
And the visualization:
Increasing amount of steps has no effect. I would assume I'm using regressor from the TensorFlow in a wrong way.
iPython notebook with the code.
It looks like your TF model does indeed work and will get there with enough steps. You need to jack it right up though - 200K showed significant improvement, almost as good as the sklearn default.
I think there are two issues:
sklearn looks like it simply solves the equation using ordinary least squares. TF's LinearRegressor uses the FtrlOptimizer. The paper indicates it is a better choice for very large datasets.
The input_fn to the model is injecting the whole training set at once, for every step. This is just a hunch, but I suspect that the FtrlOptimizer may do better if it sees batches at a time.
Instead of just changing the number of steps up a couple orders of magnitude, you can also jack the learning rate up on the optimizer (the default is 0.2) and get similarly good results from only 4k steps:
linear_regressor = tf.contrib.learn.LinearRegressor(
feature_columns=features,
optimizer=tf.train.FtrlOptimizer(learning_rate=5.0))
I met a similar problem. The solution is to check if your input_fn has enough epoch. The training maybe not converge before iterating over the whole training data several times.