Deployable probabilistic programming

Tolpin, David

doi:10.1145/3359591.3359727

Cited by 6 publications

(6 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Figure 2 shows the posterior distributions for 𝑝 inferred on the hierarchical model (Figure 2a) and through 𝑁 + 1 applications of stump-and-fungus (similar to leave-one-out cross-validation). The Infergo [Tolpin 2019] source code of the model is provided in the appendix. The inference was performed with HMC [Neal 2011] on the hierarchical model (4) and stochastic gradient HMC ] on the stump-and-fungus model (5).…”

Section: Case Study: Tumor Incidence In Ratsmentioning

confidence: 99%

How To Train Your Program: a Probabilistic Programming Pattern for Bayesian Learning From Data

Tolpin¹

2021

Preprint

Self Cite

View full text Add to dashboard Cite

We present a Bayesian approach to machine learning with probabilistic programs. In our approach, training on available data is implemented as inference on a hierarchical model. The posterior distribution of model parameters is then used to stochastically condition a complementary model, such that inference on new data yields the same posterior distribution of latent parameters corresponding to the new data as inference on a hierachical model on the combination of both previously available and new data, at a lower computation cost. We frame the approach as a design pattern of probabilistic programming referred to herein as 'stump and fungus', and illustrate realization of the pattern on a didactic case study.

show abstract

Section: Case Study: Tumor Incidence In Ratsmentioning

confidence: 99%

How To Train Your Program: a Probabilistic Programming Pattern for Bayesian Learning From Data

Tolpin¹

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…We implemented inference in stochastically differentiable probabilistic programs using Infergo [21], a probabilistic programming facility for the Go programming language. Due to reliance of Infergo on pure Go code for probabilistic programs, no changes were necessary to the way probabilistic programs are represented.…”

Section: Case Studiesmentioning

confidence: 99%

“…Some probabilistic programming frameworks require that the trace shape be specified upfront, i.e. static [2,21], while others allow introduction of trace components dynamically [8,16,7,5,22], in the course of a program execution.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Stochastically Differentiable Probabilistic Programs

Tolpin,

Zhou,

Yang

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Probabilistic programs with mixed support (both continuous and discrete latent random variables) commonly appear in many probabilistic programming systems (PPSs). However, the existence of the discrete random variables prohibits many basic gradient-based inference engines, which makes the inference procedure on such models particularly challenging. Existing PPSs either require the user to manually marginalize out the discrete variables or to perform a composing inference by running inference separately on discrete and continuous variables. The former is infeasible in most cases whereas the latter has some fundamental shortcomings. We present a novel approach to run inference efficiently and robustly in such programs using stochastic gradient Markov Chain Monte Carlo family of algorithms. We compare our stochastic gradient-based inference algorithm against conventional baselines in several important cases of probabilistic programs with mixed support, and demonstrate that it outperforms existing composing inference baselines and works almost as well as inference in marginalized versions of the programs, but with less programming effort and at a lower computation cost.

show abstract

“…An execution is summarized as an instantiation of the program trace, and the probability is a function of the trace. Some probabilistic programming frameworks require that the trace shape is specified upfront [2,10], others allow introduction of trace components dynamically [4,[6][7][8]11], in the course of a program execution.…”

Section: Introductionmentioning

confidence: 99%

Stochastic Probabilistic Programs

Tolpin,

Dobkin

2020

Preprint

Self Cite

View full text Add to dashboard Cite

We introduce the notion of a stochastic probabilistic program and present a reference implementation of a probabilistic programming facility supporting specification of stochastic probabilistic programs and inference in them. Stochastic probabilistic programs allow straightforward specification and efficient inference in models with nuisance parameters, noise, and nondeterminism. We give several examples of stochastic probabilistic programs, and compare the programs with corresponding deterministic probabilistic programs in terms of model specification and inference. We conclude with discussion of open research topics and related work.

show abstract

Deployable probabilistic programming

Cited by 6 publications

References 32 publications

How To Train Your Program: a Probabilistic Programming Pattern for Bayesian Learning From Data

How To Train Your Program: a Probabilistic Programming Pattern for Bayesian Learning From Data

Stochastically Differentiable Probabilistic Programs

Stochastic Probabilistic Programs

Contact Info

Product

Resources

About