A design proposal for Gen: probabilistic programming with fast custom inference via code generation

Probabilistic programming refers to the idea of using standard programming constructs for specifying probabilistic models from machine learning and statistics, and employing generic inference algorithms for answering various queries on these models, such as posterior inference and estimation of model evidence. Although this idea itself is not new and was, in fact, explored by several programming-language and statistics researchers in the early 2000, it is only in the last few years that probabilistic programming has gained a large amount of attention among researchers in machine learning and programming languages, and that expressive and efficient probabilistic programming systems (such as Anglican, Church, Figaro, Infer.net, PSI, PyMC, Stan, and Venture) started to appear and have been taken up by a nontrivial number of users. The primary goal of my talk is to introduce probabilistic programming to the CONCUR/QUEST/FORMATS audience. At the end of my talk, I want the audience to understand basic results and techniques in probabilistic programming and to feel that these results and techniques are relevant or at least related to what she or he studies, although they typically come from foreign research areas, such as machine learning and statistics. My talk will contain both technical materials and lessons that I learnt from my machine-learning colleagues in Oxford, who are developing a highly-expressive higher-order probabilistic programming language, called Anglican. It will also include my work on the denotational semantics of higher-order probabilistic programming languages and their inference algorithms, which are jointly pursued with colleagues in Cambridge, Edinburgh, Oxford and Tubingen.

Probabilistic programming languages (PPLs) are an expressive means for creating and reasoning about probabilistic models. Unfortunately hybrid probabilistic programs that involve both continuous and discrete structures are not well supported by today’s PPLs. In this paper we develop a new approximate inference algorithm for hybrid probabilistic programs that first discretizes the continuous distributions and then performs discrete inference on the resulting program. The key novelty is a form of discretization that we call bit blasting , which uses a binary representation of numbers such that a domain of 2 b discretized points can be succinctly represented as a discrete probabilistic program over poly b Boolean random variables. Surprisingly, we prove that many common continuous distributions can be bit blasted in a manner that incurs no loss of accuracy over an explicit discretization and supports efficient probabilistic inference. We have built a probabilistic programming system for hybrid programs called HyBit , which employs bit blasting followed by discrete probabilistic inference. We empirically demonstrate the benefits of our approach over existing sampling-based and symbolic inference approaches

Probabilistic Logic Programming extends Logic Programming by enabling the representation of uncertain information by means of probability theory. Probabilistic Logic Programming is at the intersection of two wider research fields: the integration of logic and probability and Probabilistic Programming. Logic enables the representation of complex relations among entities while probability theory is useful for model uncertainty over attributes and relations. Combining the two is a very active field of study. Probabilistic Programming extends programming languages with probabilistic primitives that can be used to write complex probabilistic models. Algorithms for the inference and learning tasks are then provided automatically by the system. Probabilistic Logic programming is at the same time a logic language, with its knowledge representation capabilities, and a Turing complete language, with its computation capabilities, thus providing the best of both worlds. Since its birth, the field of Probabilistic Logic Programming has seen a steady increase of activity, with many proposals for languages and algorithms for inference and learning. Foundations of Probabilistic Logic Programming aims at providing an overview of the field with a special emphasis on languages under the Distribution Semantics, one of the most influential approaches. The book presents the main ideas for semantics, inference, and learning and highlights connections between the methods. Many examples of the book include a link to a page of the web application http://cplint.eu where the code can be run online.

Probabilistic programming languages aim to separate the user-specified model from the inference runtime system, allowing users to focus on model design and leave inference to the compiler and runtime systems. However, the structure of the model can affect the inference efficiency, something that is often utilized when implementing Bayesian inference solutions without probabilistic programming. In this paper, we consider a specific class of important models within phylogenetics, which are used for tree inference. Algorithms utilizing the structures of trees in such models have been used for a long time within phylogenetics, outside the scope of probabilistic programming. In particular, the use of belief propagation can significantly speed up probabilistic tree inference when combined with other approximate inferences, such as MCMC. However, there is an open problem of how to incorporate both belief-propagation (marginalization) of a subset of latent variables in typed universal probabilistic programming languages and still be able to apply approximate inference for the rest of the variables. In this paper, we propose a solution to this problem by using an approach where a user only needs to mark which latent variables in the program should be marginalized, and the rest of the belief propagation is handled automatically. We call this approach annotated automatic pruning, acknowledging the original terminology of pruning from the phylogenetic literature. We evaluate our approach using non-trivial models for tree inference, assessing performance, and testing correctness by comparing to MrBayes as the reference implementation.

Modern probabilistic programming languages aim to formalize and automate key aspects of probabilistic modeling and inference. Many languages provide constructs for programmable inference that enable developers to improve inference speed and accuracy by tailoring an algorithm for use with a particular model or dataset. Unfortunately, it is easy to use these constructs to write unsound programs that appear to run correctly but produce incorrect results. To address this problem, we present a denotational semantics for programmable inference in higher-order probabilistic programming languages, along with a type system that ensures that well-typed inference programs are sound by construction. A central insight is that the type of a probabilistic expression can track the space of its possible execution traces, not just the type of value that it returns, as these traces are often the objects that inference algorithms manipulate. We use our semantics and type system to establish soundness properties of custom inference programs that use constructs for variational, sequential Monte Carlo, importance sampling, and Markov chain Monte Carlo inference.

Probabilistic Programming Languages (PPLs) allow users to encode statistical inference problems and automatically apply an inference algorithm to solve them. Popular inference algorithms for PPLs, such as sequential Monte Carlo (SMC) and Markov chain Monte Carlo (MCMC), are built around checkpoints—relevant events for the inference algorithm during the execution of a probabilistic program. Deciding the location of checkpoints is, in current PPLs, not done optimally. To solve this problem, we present a static analysis technique that automatically determines checkpoints in programs, relieving PPL users of this task. The analysis identifies a set of checkpoints that execute in the same order in every program run—they are aligned. We formalize alignment, prove the correctness of the analysis, and implement the analysis as part of the higher-order functional PPL Miking CorePPL. By utilizing the alignment analysis, we design two novel inference algorithm variants: aligned SMC and aligned lightweight MCMC. We show, through real-world experiments, that they significantly improve inference execution time and accuracy compared to standard PPL versions of SMC and MCMC.

Probabilistic programming languages (PPLs) allow programmers to construct statistical models and then simulate data or perform inference over them. Many PPLs restrict models to a particular instance of simulation or inference, limiting their reusability. In other PPLs, models are not readily composable. Using Haskell as the host language, we present an embedded domain specific language based on algebraic effects, where probabilistic models are modular, first-class, and reusable for both simulation and inference. We also demonstrate how simulation and inference can be expressed naturally as composable program transformations using algebraic effect handlers.

While Search-Based Software Testing (SBST) has improved significantly in the last decade we propose that more flexible, probabilistic models can be leveraged to improve it further. Rather than searching for an individual, or even sets of, test case(s) or datum(s) that fulfil specific needs the goal can be to learn a generative model tuned to output a useful family of values. Such generative models can naturally be decomposed into a structured generator and a probabilistic model that determines how to make non-deterministic choices during generation. While the former constrains the generation process to produce valid values the latter allows learning and tuning to specific goals. SBST techniques differ in their level of integration of the two but, regardless of how close it is, we argue that the flexibility and power of the probabilistic model will be a main determinant of success. In this short paper, we present how some existing SBST techniques can be viewed from this perspective and then propose additional techniques for flexible generative modelling the community should consider. In particular, Probabilistic Programming languages (PPLs) and Genetic Programming (GP) should be investigated since they allow for very flexible probabilistic modelling. Benefits could range from utilising the multiple program executions that SBST techniques typically require to allowing the encoding of high-level test strategies.

Probabilistic programming languages (PPLs) are at the interface between statistics and the theory of programming languages. PPLs formulate statistical models as stochastic programs that enable automatic inference algorithms and optimization. Pyro [1] and its sibling NumPyro [2] are PPLs built on top of the deep learning frameworks PyTorch [3] and Jax [4], respectively. Both PPLs provide simple, highly similar interfaces for inference using efficient implementations of Hamiltonian Monte Carlo (HMC), the No-U-Turn Sampler (NUTS), and Stochastic Variational Inference (SVI). They automatically generate variational distributions from a model, automatically enumerate discrete variables, and support formulating deep probabilistic models such as variational autoencoders and deep Markov models. The Sine von Mises distribution and its skewed variant are toroidal distributions relevant to protein bioinformatics. They provide a natural way to model the dihedral angles of protein structures, which is important in protein structure prediction, simulation and analysis. We present efficient implementations of the Sine von Mises distribution and its skewing in Pyro and NumPyro, and devise a simulation method that increases efficiency with several orders of magnitude when using parallel hardware (i.e., modern CPUs, GPUs, and TPUs). We demonstrate the use of the skewed Sine von Mises distribution by modeling dihedral angles of proteins using a Bayesian mixture model inferred using NUTS, exploiting NumPyro's facilities for automatic enumeration [5].

This system description paper introduces the probabilistic programming language Hakaru10, for expressing, and performing inference on (general) graphical models. The language supports discrete and continuous distributions, mixture distributions and conditioning. Hakaru10 is a DSL embedded in Haskell and supports Monte-Carlo Markov Chain (MCMC) inference.

Opening doors is an essential task that a robot should perform. In this paper, we propose a logical approach to predict the action of opening doors, together with the action point where the action should be performed. The input of our system is a pair of bounding boxes of the door and door handle, together with background knowledge in the form of logical rules. Learning and inference are performed with the probabilistic programming language ProbLog. We evaluate our approach on a doors dataset and we obtain encouraging results. Additionally, a comparison to a propositional decision tree shows the benefits of using a probabilistic programming language such as ProbLog.

Gaussian processes are powerful non-parametric probabilistic models for stochastic functions. However, the direct implementation entails a complexity that is computationally intractable when the number of observations is large, especially when estimated with fully Bayesian methods such as Markov chain Monte Carlo. In this paper, we focus on a low-rank approximate Bayesian Gaussian processes, based on a basis function approximation via Laplace eigenfunctions for stationary covariance functions. The main contribution of this paper is a detailed analysis of the performance, and practical recommendations for how to select the number of basis functions and the boundary factor. Intuitive visualizations and recommendations, make it easier for users to improve approximation accuracy and computational performance. We also propose diagnostics for checking that the number of basis functions and the boundary factor are adequate given the data. The approach is simple and exhibits an attractive computational complexity due to its linear structure, and it is easy to implement in probabilistic programming frameworks. Several illustrative examples of the performance and applicability of the method in the probabilistic programming language Stan are presented together with the underlying Stan model code.

Popular application domains such as multimedia processing, machine learning, and big-data analytics operate on inherently noisy data and make decisions under uncertainty. While these applications are often good candidates for both algorithmic and system-level approximation, a major open challenge is how to analyze the influence of noisy data and candidate approximations on the application's outputs.At the same time, probabilistic programming languages provide an intuitive way to model uncertainty by expressing complex probabilistic models as computer programs. The talk will give an overview of PSI (http://www.psisolver.org), a system for exact symbolic inference. PSI computes succinct symbolic representations of the joint posterior distribution represented by a probabilistic program using static analysis. PSI supports programs with both discrete and continuous distributions. It can compute answers to various posterior distribution queries, expectation queries and assertion queries using its own back-end for symbolic reasoning. This talk will present how we can represent some problems in approximate computing as probabilistic programs and use PSI to automatically get symbolic expressions that represent the distributions of the output error.

A critical task for system operators is the precise identification of the root causes underlying an error situation. This identification is fundamental in deciding optimal maintenance actions, such as replacing a component versus calibrating it. However, the actual causes of an error are often neither measured nor unique. The measured quantities are the result of complex interactions between different error causes and system variables. Root cause identification in this context becomes a matter of inferring hidden causes from their measurable effects. This challenge is notably pronounced in cyber-physical systems comprising control loops. Control mechanisms, integral to maintaining system performance, introduce a layer of complexity in diagnostics and ultimately complicate the isolation of the underlying causes of errors. To address this challenge, we introduce a two-step approach to derive the hidden causes as a statistical inference task. First, we develop a generative model leveraging existing control software and expert-based insights into the mechanisms of errors, i.e., a simulator of synthetic data given some hidden error causes. Then, we transform the generative model into a probabilistic program on which statistical inference can be executed within a probabilistic programming language framework. This inference effectively estimates the hidden causes given some measured data from the system. Being intrinsically a statistical approach, these inferences come with a confidence interval. We applied this methodology to an industrial printer’s sheet transport belt, operating in a closed-loop configuration. Our approach successfully discerned the contributions of three distinct hidden causes to the belt’s deviation from its intended position. This paper highlights the efficacy of generative modeling followed by a probabilistic programming approach in unraveling complex interactions within cyber-physical systems for optimal maintenance.

Satellite data provides essential insights into the spatiotemporal distribution of CO2 concentrations. However, many atmospheric inverse models fail to adequately incorporate the spatial and temporal correlations inherent in satellite observations and often lack rigorous methods for estimating parameters like spatial length scales. We introduce an inference model that processes the spatiotemporal covariance in satellite data and estimates hyperparameters such as covariance length scales. Our approach uses the Gaussian process (GP) machine learning (ML) and modern probabilistic programming languages (PPLs) to perform atmospheric inversions of emissions from satellite data. We develop a GP ML inversion system based on modern PPLs and the GEOS-Chem chemical transport model, simulating atmospheric CO2 concentrations corresponding to the Orbiting Carbon Observatory-2/3 (OCO-2/3) data for July 2020. In our supervised learning framework, we treat the GEOS-Chem simulated data set as the target, with predictors derived by scaling the target with sector-specific factors hidden from the GP machine. Our results show that the GP model, combined with GPU-enabled PPLs, effectively retrieves true emission scaling factors and infers noise levels concealed within the data. This suggests that our method could be applied over larger areas with more complex covariance structures, enabling comprehensive analysis of the spatiotemporal patterns observed in OCO-2/3 and similar satellite data sets.