probability
probability copied to clipboard
tensorflow
Probabilistic reasoning and statistical analysis in TensorFlow
[TFP is awesome, thanks so much for all the amazing functionality!] I've been reading the copulas primer at https://www.tensorflow.org/probability/examples/Gaussian_Copula Twice in the first couple of paragraphs it mentions that a...
Dear maintainers, I was wondering if there is any way of extending tfd.HiddenMarkovModel to a hidden semi-Markov setting, where emissions depend not only on the last seen hidden state but...
Refactored particle_filter.py script Added core function sequential_monte_carlo in particle_filter which works on WeightedParticles instead of distributions.
What are the equivalent implementation of the following function from `scikit-learn` about [Random Projection?](https://scikit-learn.org/stable/modules/random_projection.html#) ```python from sklearn import random_projection random_projection.johnson_lindenstrauss_min_dim random_projection.GaussianRandomProjection random_projection.SparseRandomProjection random_projection.SparseRandomProjection(compute_inverse_components=True) ``` And how that can be used...
From here: https://groups.google.com/a/tensorflow.org/g/tfprobability/c/_9EuU3FnErY/m/cpNQENy6AQAJ > 1) Re: saving models: I think it'd be both advantageous and straightforward for us to implement __setstate__ and __getstate__ in Distributions, Bijectors, and DeferredTensor. For Distributions...
I'd like to extend the hidden Markov model implementation so that it can handle transition matrices that are very sparse. For my usecase, I'd like to extend it such that...
We would be interested to have the Faddeeva function available (erfcx for complex arguments) and would like to implement and then contribute it (as we will have a GSoC student)....
Would it be feasible at all to add PyTorch support as a substrate? Besides the obvious use case, this would enable researchers to implement framework-agnostic probabilistic algorithms that use a...
Are there any plans to add a Zero-Inflated Poisson (ZIP) and Zero-Inflated Negative Binomial (ZINB) to TFP? Those are usually very common distributions in other packages, and it shouldn't be...
When doing the model prediction on test data, ``` inputs = Input(shape=(X_train.shape[1],)) rate = Dense(1, activation=tf.exp)(inputs) p_y = tfp.layers.DistributionLambda(tfd.Poisson)(rate) model_p = Model(inputs=inputs, outputs=p_y) def NLL(y_true, y_hat): return -y_hat.log_prob(y_true) model_p.compile(Adam(learning_rate=0.001), loss=NLL)...
Metadata
Owner
Metadata
Probabilistic reasoning and statistical analysis in TensorFlow