Inverse Ising Problem Research Articles

Accelerating convergence of inference in the inverse Ising problem

Inferring modelling parameters of dynamical processes from observational data is an important inverse problem in statistical physics. In this paper, instead of passively observing the dynamics for inference, we focus on strategically manipulating dynamics to generate data that gives more accurate estimators within fewer observations. For this purpose, we consider the inference problem rooted in the Ising model with two opposite external fields, assuming that the strength distribution of one of the fields (labelled as passive) is unknown and needs to be inferred. In contrast, the other field (labelled active) is strategically deployed to interact with the Ising dynamics in such a way as to improve the accuracy of estimates of inferring the opposing passive field. By comparing to benchmark cases, we first demonstrate that it is possible to accelerate the inference by strategically interacting with the Ising dynamics. We then apply series expansions to obtain an approximation of the optimized influence configurations in the high-temperature region. Furthermore, by using mean-field estimates, we also demonstrate the applicability of the method in a more general scenario where real-time tracking of the system is infeasible. Last, analysing the optimized influence profiles, we describe heuristics for manipulating the Ising dynamics for faster inference. For example, we show that agents targeted more strongly by the passive field should also be strongly targeted by the active one.

Interaction graph learning of line cascading failure in power networks and its statistical properties

We consider line failure cascading in power networks where an initial random failure of a few lines leads to consecutive other line overloads and failures before the system settles in a steady state. Such cascades are rooted in non-obvious, long-range, and higher-order couplings among the lines’ flows induced by physical constraints on the network. Failure interaction graph encodes which and to what extent other lines in a networked system are affected after each line failure and can help to predict the final state after an initial disturbance. We perform data analytics on the final lines’ steady states of cascade trajectories to infer a specific line’s state given the states of others. We use a generative model to reconstruct possible steady states, and a predictive model aims to predict the probability of each line’s failures after the initial failure as a regression problem. The generative model uses regularized pseudolikelihood estimator to infer interaction weights by solving the inverse Ising problem and deploys Glauber dynamics to generate steady states. The discriminative model uses boosted trees to efficiently learn over training and predict over test data the state of each line as a target finding an appropriate subset of other lines’ states as explanatory variables. We analyze the degree distribution of the corresponding interaction graphs to study the number of other components affected by each line failure (out-degree) or the number of lines that affect the state of a given line (in-degree). Both models show that the in-degree follows a power-law distribution. Finally, we discuss the possible application of the interaction graph for early link removal to mitigate the failure-cascading consequences.

Composite Spatial Monte Carlo Integration Based on Generalized Least Squares

Although evaluation of the expectations on the Ising model is essential in various applications, it is mostly infeasible because of intractable multiple summations. Spatial Monte Carlo integration (SMCI) is a sampling-based approximation. It can provide high-accuracy estimations for such intractable expectations. To evaluate the expectation of a function of variables in a specific region (called target region), SMCI considers a larger region containing the target region (called sum region). In SMCI, the multiple summation for the variables in the sum region is precisely executed, and that in the outer region is evaluated by the sampling approximation such as the standard Monte Carlo integration. It is guaranteed that the accuracy of the SMCI estimator improves monotonically as the size of the sum region increases. However, a haphazard expansion of the sum region could cause a combinatorial explosion. Therefore, we hope to improve the accuracy without such an expansion. In this paper, based on the theory of generalized least squares (GLS), a new effective method is proposed by combining multiple SMCI estimators. The validity of the proposed method is demonstrated theoretically and numerically. The results indicate that the proposed method can be effective in the inverse Ising problem (or Boltzmann machine learning).

A noniterative solution to the inverse Ising problem using a convex upper bound on the partition function

The inverse Ising problem, or the learning of Ising models, is notoriously difficult, as evaluating the partition function has a large computational cost. To quickly solve this problem, inverse formulas using approximation methods such as the Bethe approximation have been developed. In this paper, we employ the tree-reweighted (TRW) approximation to construct a new inverse formula. An advantage of using the TRW approximation is that it provides a rigorous upper bound on the partition function, allowing us to optimize a lower bound for the learning objective function. We show that the moment-matching and self-consistency conditions can be solved analytically, and we obtain an analytic form of the approximate interaction matrix as a function of the given data statistics. Using this solution, we can compute the interaction matrix that is optimal to the approximate objective function without iterative computation. To evaluate the accuracy of the derived learning formula, we compared our formula to those obtained by other approximations. From our experiments on reconstructing interaction matrices, we found that the proposed formula gives the best estimates in models with strongly attractive interactions on various graphs.

Structure learning in inverse Ising problems using ℓ 2-regularized linear estimator

The inference performance of the pseudolikelihood method is discussed in the framework of the inverse Ising problem when the ℓ 2-regularized (ridge) linear regression is adopted. This setup is introduced for theoretically investigating the situation where the data generation model is different from the inference one, namely the model mismatch situation. In the teacher-student scenario under the assumption that the teacher couplings are sparse, the analysis is conducted using the replica and cavity methods, with a special focus on whether the presence/absence of teacher couplings is correctly inferred or not. The result indicates that despite the model mismatch, one can perfectly identify the network structure using naive linear regression without regularization when the number of spins N is smaller than the dataset size M, in the thermodynamic limit N → ∞. Further, to access the underdetermined region M < N, we examine the effect of the ℓ 2 regularization, and find that biases appear in all the coupling estimates, preventing the perfect identification of the network structure. We, however, find that the biases are shown to decay exponentially fast as the distance from the center spin chosen in the pseudolikelihood method grows. Based on this finding, we propose a two-stage estimator: in the first stage, the ridge regression is used and the estimates are pruned by a relatively small threshold; in the second stage the naive linear regression is conducted only on the remaining couplings, and the resultant estimates are again pruned by another relatively large threshold. This estimator with the appropriate regularization coefficient and thresholds is shown to achieve the perfect identification of the network structure even in 0 < M/N < 1. Results of extensive numerical experiments support these findings.

Detecting Many-Body Bell Nonlocality by Solving Ising Models.

Bell nonlocality represents the ultimate consequence of quantum entanglement, fundamentally undermining the classical tenet that spatially separated degrees of freedom possess objective attributes independently of the act of their measurement. Despite its importance, probing Bell nonlocality in many-body systems is considered to be a formidable challenge, with a computational cost scaling exponentially with system size. Here we propose and validate an efficient variational scheme, based on the solution of inverse classical Ising problems, which in polynomial time can probe whether an arbitrary set of quantum data is compatible with a local theory; and, if not, it delivers the many-body Bell inequality most strongly violated by the quantum data. We use our approach to unveil new many-body Bell inequalities, violated by suitable measurements on paradigmatic quantum states (the low-energy states of Heisenberg antiferromagnets), paving the way to systematic Bell tests in the many-body realm.

A density consistency approach to the inverse Ising problem

We propose a novel approach to the inverse Ising problem which employs the recently introduced density consistency approximation (DC) to determine the model parameters (couplings and external fields) maximizing the likelihood of given empirical data. This method allows for closed-form expressions of the inferred parameters as a function of the first and second empirical moments. Such expressions have a similar structure to the small-correlation expansion derived in reference Sessak and Monasson (2009 J. Phys. A: Math. Theor. 42 055001), of which they provide an improvement in the case of non-zero magnetization at low temperatures, as well as in presence of random external fields. The present work provides an extensive comparison with most common inference methods used to reconstruct the model parameters in several regimes, i.e. by varying both the network topology and the distribution of fields and couplings. The comparison shows that no method is uniformly better than every other one, but DC appears nevertheless as one of the most accurate and reliable approaches to infer couplings and fields from first and second moments in a significant range of parameters.

Learning performance in inverse Ising problems with sparse teacher couplings

We investigate the learning performance of the pseudolikelihood maximization method for inverse Ising problems. In the teacher–student scenario under the assumption that the teacher’s couplings are sparse and the student does not know the graphical structure, the learning curve and order parameters are assessed in the typical case using the replica and cavity methods from statistical mechanics. Our formulation is also applicable to a certain class of cost functions having locality; the standard likelihood does not belong to that class. The derived analytical formulas indicate that the perfect inference of the presence/absence of the teacher’s couplings is possible in the thermodynamic limit taking the number of spins N as infinity while keeping the dataset size M proportional to N, as long as α = M/N > 2. Meanwhile, the formulas also show that the estimated coupling values corresponding to the truly existing ones in the teacher tend to be overestimated in the absolute value, manifesting the presence of estimation bias. These results are considered to be exact in the thermodynamic limit on locally tree-like networks, such as the regular random or Erdős–Rényi graphs. Numerical simulation results fully support the theoretical predictions. Additional biases in the estimators on loopy graphs are also discussed.

Mi3-GPU: MCMC-based inverse Ising inference on GPUs for protein covariation analysis

Inverse Ising inference is a method for inferring the coupling parameters of a Potts/Ising model based on observed site-covariation, which has found important applications in protein physics for detecting interactions between residues in protein families. We introduce Mi3-GPU (“mee-three”, for MCMC Inverse Ising Inference) software for solving the inverse Ising problem for protein-sequence datasets with few analytic approximations, by parallel Markov-Chain Monte Carlo sampling on GPUs. We also provide tools for analysis and preparation of protein-family Multiple Sequence Alignments (MSAs) to account for finite-sampling issues, which are a major source of error or bias in inverse Ising inference. Our method is “generative” in the sense that the inferred model can be used to generate synthetic MSAs whose mutational statistics (marginals) can be verified to match the dataset MSA statistics up to the limits imposed by the effects of finite sampling. Our GPU implementation enables the construction of models which reproduce the covariation patterns of the observed MSA with a precision that is not possible with more approximate methods. The main components of our method are a GPU-optimized algorithm to greatly accelerate MCMC sampling, combined with a multi-step Quasi-Newton parameter-update scheme using a “Zwanzig reweighting” technique. We demonstrate the ability of this software to produce generative models on typical protein family datasets for sequence lengths L∼300 with 21 residue types with tens of millions of inferred parameters in short running times. Program summaryProgram Title: Mi3-GPUProgram Files doi:http://dx.doi.org/10.17632/ftbcfy2p35.1Licensing provisions: GPLv3Programming languages: Python3, OpenCL, CNature of problem: Mi3-GPU solves the inverse Ising problem for application in protein covariation analysis. The goal is to infer “coupling” parameters between positions in a Multiple Sequence Alignment of a protein family, with many applications including protein-contact prediction and fitness prediction.Solution method: Mi3-GPU solves the inverse Ising problem with few approximations using Markov-Chain Monte Carlo methods with Quasi-Newton optimization on GPUs. This problem previously has been approached by more approximate methods using analytic approximations including “message Passing”, “Susceptibility Propagation”, “mean-field” methods, pseudolikelihood approximations, and cluster expansion. The software leverages GPU to accelerate MCMC sampling and a histogram reweighting technique to accelerate parameter optimization.

Applications of the inverse Ising problem to biological data analysis

Applications of the inverse Ising problem to biological data analysis

Adaptive cluster expansion for Ising spin models

We propose an algorithm to obtain numerically approximate solutions of the direct Ising problem, that is, to compute the free energy and the equilibrium observables of spin systems with arbitrary two-spin interactions. To this purpose we use the Adaptive Cluster Expansion method, originally developed to solve the inverse Ising problem, that is, to infer the interactions from the equilibrium correlations. The method consists in iteratively constructing and selecting clusters of spins, computing their contributions to the free energy and discarding clusters whose contribution is lower than a fixed threshold. The properties of the cluster expansion and its performance are studied in detail on one dimensional, two dimensional, random and fully connected graphs with homogeneous or heterogeneous fields and couplings. We discuss the differences between different representations (Boolean and Ising) of the spin variables.

Reconstructing positive and negative couplings in Ising spin networks by sorted local transfer entropy.

We employ the sorted local transfer entropy (SLTE) to reconstruct the coupling strengths of Ising spin networks with positive and negative couplings (J_{ij}), using only the time-series data of the spins. The SLTE method is model-free in the sense that no knowledge of the underlying dynamics of the spin system is required and is applicable to a broad class of systems. Contrary to the inference of coupling from pairwise transfer entropy, our method can reliably distinguish spin pair interactions with positive and negative couplings. The method is tested for the inverse Ising problem for different J_{ij} distributions and various spin dynamics, including synchronous and asynchronous update Glauber dynamics and Kawasaki exchange dynamics. It is found that the pairwise SLTE is proportional to the pairwise coupling strength to a good extent for all cases studied. In addition, the reconstruction works well for both the equilibrium and nonequilibrium cases of the time-series data. Comparison to other inverse Ising problem approaches using mean-field equations is also discussed.

Market basket analysis by solving the inverse Ising problem: Discovering pairwise interaction strengths among products

Large datasets containing the purchasing information of thousands of consumers are difficult to analyze because the possible number of different combinations of products is huge. Thus, market baskets analysis to obtain useful information and find interesting pattern of buying behavior could be a daunting task. Based on the maximum entropy principle, we build a probabilistic model that explains the probability of occurrence of market baskets which is equivalent to Ising models. This type of model allows us to understand and to explore the functional interactions among products that make up the market offer. Additionally, the parameters of the model inferred using Boltzmann learning, allow us to suggest that the buying behavior is very similar to the spin-glass physical system. Moreover, we show that the resulting parameters of the model could be useful to describe the hierarchical structure of the system which leads to interesting information about the different market baskets.

HoDCA: higher order direct-coupling analysis

BackgroundDirect-coupling analysis (DCA) is a method for protein contact prediction from sequence information alone. Its underlying principle is parameter estimation for a Hamiltonian interaction function stemming from a maximum entropy model with one- and two-point interactions. Vastly growing sequence databases enable the construction of large multiple sequence alignments (MSA). Thus, enough data exists to include higher order terms, such as three-body correlations.ResultsWe present an implementation of hoDCA, which is an extension of DCA by including three-body interactions into the inverse Ising problem posed by parameter estimation. In a previous study, these three-body-interactions improved contact prediction accuracy for the PSICOV benchmark dataset. Our implementation can be executed in parallel, which results in fast runtimes and makes it suitable for large-scale application.ConclusionOur hoDCA software allows improved contact prediction using the Julia language, leveraging power of multi-core machines in an automated fashion.

Inverse Ising problem in continuous time: A latent variable approach.

We consider the inverse Ising problem: the inference of network couplings from observed spin trajectories for a model with continuous time Glauber dynamics. By introducing two sets of auxiliary latent random variables we render the likelihood into a form which allows for simple iterative inference algorithms with analytical updates. The variables are (1) Poisson variables to linearize an exponential term which is typical for point process likelihoods and (2) Pólya-Gamma variables, which make the likelihood quadratic in the coupling parameters. Using the augmented likelihood, we derive an expectation-maximization (EM) algorithm to obtain the maximum likelihood estimate of network parameters. Using a third set of latent variables we extend the EM algorithm to sparse couplings via L1 regularization. Finally, we develop an efficient approximate Bayesian inference algorithm using a variational approach. We demonstrate the performance of our algorithms on data simulated from an Ising model. For data which are simulated from a more biologically plausible network with spiking neurons, we show that the Ising model captures well the low order statistics of the data and how the Ising couplings are related to the underlying synaptic structure of the simulated network.

Statistical mechanics of the inverse Ising problem and the optimal objective function

The inverse Ising problem seeks to reconstruct the parameters of an Ising Hamiltonian on the basis of spin configurations sampled from the Boltzmann measure. Over the last decade, many applications of the inverse Ising problem have arisen, driven by the advent of large-scale data across different scientific disciplines. Recently, strategies to solve the inverse Ising problem based on convex optimisation have proven to be very successful. These approaches maximise particular objective functions with respect to the model parameters. Examples are the pseudolikelihood method and interaction screening. In this paper, we establish a link between approaches to the inverse Ising problem based on convex optimisation and the statistical physics of disordered systems. We characterise the performance of an arbitrary objective function and calculate the objective function which optimally reconstructs the model parameters. We evaluate the optimal objective function within a replica-symmetric ansatz and compare the results of the optimal objective function with other reconstruction methods. Apart from giving a theoretical underpinning to solving the inverse Ising problem by convex optimisation, the optimal objective function outperforms state-of-the-art methods, albeit by a small margin.

Inverse statistical problems: from the inverse Ising problem to data science

Inverse problems in statistical physics are motivated by the challenges of `big data' in different fields, in particular high-throughput experiments in biology. In inverse problems, the usual procedure of statistical physics needs to be reversed: Instead of calculating observables on the basis of model parameters, we seek to infer parameters of a model based on observations. In this review, we focus on the inverse Ising problem and closely related problems, namely how to infer the coupling strengths between spins given observed spin correlations, magnetisations, or other data. We review applications of the inverse Ising problem, including the reconstruction of neural connections, protein structure determination, and the inference of gene regulatory networks. For the inverse Ising problem in equilibrium, a number of controlled and uncontrolled approximate solutions have been developed in the statistical mechanics community. A particularly strong method, pseudolikelihood, stems from statistics. We also review the inverse Ising problem in the non-equilibrium case, where the model parameters must be reconstructed based on non-equilibrium statistics.

A statistical physics approach to learning curves for the inverse Ising problem

Using methods of statistical physics, we analyse the error of learning couplings in large Ising models from independent data (the inverse Ising problem). We concentrate on learning based on local cost functions, such as the pseudo-likelihood method for which the couplings are inferred independently for each spin. Assuming that the data are generated from a true Ising model, we compute the reconstruction error of the couplings using a combination of the replica method with the cavity approach for densely connected systems. We show that an explicit estimator based on a quadratic cost function achieves minimal reconstruction error, but requires the length of the true coupling vector as prior knowledge. A simple mean field estimator of the couplings which does not need such knowledge is asymptotically optimal, i.e. when the number of observations is much larger than the number of spins. Comparison of the theory with numerical simulations shows excellent agreement for data generated from two models with random couplings in the high temperature region: a model with independent couplings (Sherrington–Kirkpatrick model), and a model where the matrix of couplings has a Wishart distribution.

Detecting multi-spin interactions in the inverse Ising problem

While the usual goal in Monte Carlo (MC) simulations of Ising models is the efficient generation of spin configurations with Boltzmann probabilities, the inverse problem is to determine the coupling constants from a given set of spin configurations. Most recent work has been limited to local magnetic fields and pair-wise interactions. We have extended solutions to multi-spin interactions, using correlation function matching (CFM). A more serious limitation of previous work has been the uncertainty of whether a chosen set of interactions is capable of faithfully representing real data. We show how our confirmation testing method uses an additional MC simulation to detect significant interactions that might be missing in the assumed representation of the data.

Improving landscape inference by integrating heterogeneous data in the inverse Ising problem.

The inverse Ising problem and its generalizations to Potts and continuous spin models have recently attracted much attention thanks to their successful applications in the statistical modeling of biological data. In the standard setting, the parameters of an Ising model (couplings and fields) are inferred using a sample of equilibrium configurations drawn from the Boltzmann distribution. However, in the context of biological applications, quantitative information for a limited number of microscopic spins configurations has recently become available. In this paper, we extend the usual setting of the inverse Ising model by developing an integrative approach combining the equilibrium sample with (possibly noisy) measurements of the energy performed for a number of arbitrary configurations. Using simulated data, we show that our integrative approach outperforms standard inference based only on the equilibrium sample or the energy measurements, including error correction of noisy energy measurements. As a biological proof-of-concept application, we show that mutational fitness landscapes in proteins can be better described when combining evolutionary sequence data with complementary structural information about mutant sequences.