We introduce a conditional compression problem and propose a fast framework for tackling it. The problem is how to quickly compress a pretrained large neural network into optimal smaller networks given target contexts, e.g. a context involving only a subset of classes or a context where only limited compute resource is available. To solve this, we propose an efficient Bayesian framework to compress a given large network into much smaller size tailored to meet each contextual requirement. We employ a hypernetwork to parameterize the posterior distribution of weights given conditional inputs and minimize a variational objective of this Bayesian neural network. To further reduce the network sizes, we propose a new input-output group sparsity factorization of weights to encourage more sparseness in the generated weights. Our methods can quickly generate compressed networks with significantly smaller sizes than baseline methods.

This is an interesting paper that distils structure learning in Bayesian networks (BNs) and kernel methods in a quest to produce more flexible distributional assumptions. Conditional (linear) Gaussian Bayesian networks (CGBNs) have been well explored in the literature for some time, to the point that they now appear in many recent textbooks [1-3]. The authors address one of the key limitations of CGBNS, that they can only capture linear dependencies between the continuous variables they contain, and remove it by replacing (mixtures of) linear regression models with more general kernel densities. Dependencies between discrete variables were already flexible, because the conditional probability tables that parametrise them essentially act as a saturated model[4]. It is not obvious that more flexibility will produce better models for whatever task we have in mind: it can also lead to overfitting, instability and hyperparameter tuning problems.

This paper describes heavy-tailed extensions of a state-of-the-art versatile blind source separation method called fast multichannel nonnegative matrix factorization (FastMNMF) from a unified point of view. The common way of deriving such an extension is to replace the multivariate complex Gaussian distribution in the likelihood function with its heavy-tailed generalization, e.g., the multivariate complex Student's t and leptokurtic generalized Gaussian distributions, and tailor-make the corresponding parameter optimization algorithm. Using a wider class of heavy-tailed distributions called a Gaussian scale mixture (GSM), i.e., a mixture of Gaussian distributions whose variances are perturbed by positive random scalars called impulse variables, we propose GSM-FastMNMF and develop an expectationmaximization algorithm that works even when the probability density function of the impulse variables have no analytical expressions. We show that existing heavy-tailed FastMNMF extensions are instances of GSM-FastMNMF and derive a new instance based on the generalized hyperbolic distribution that include the normal-inverse Gaussian, Student's t, and Gaussian distributions as the special cases. Our experiments show that the normalinverse Gaussian FastMNMF outperforms the state-of-the-art FastMNMF extensions and ILRMA model in speech enhancement and separation in terms of the signal-to-distortion ratio.

Catastrophe (CAT) bond markets are incomplete and hence carry uncertainty in instrument pricing. As such various pricing approaches have been proposed, but none treat the uncertainty in catastrophe occurrences and interest rates in a sufficiently flexible and statistically reliable way within a unifying asset pricing framework. Consequently, little is known empirically about the expected risk-premia of CAT bonds. The primary contribution of this paper is to present a unified Bayesian CAT bond pricing framework based on uncertainty quantification of catastrophes and interest rates. Our framework allows for complex beliefs about catastrophe risks to capture the distinct and common patterns in catastrophe occurrences, and when combined with stochastic interest rates, yields a unified asset pricing approach with informative expected risk premia. Specifically, using a modified collective risk model -- Dirichlet Prior-Hierarchical Bayesian Collective Risk Model (DP-HBCRM) framework -- we model catastrophe risk via a model-based clustering approach. Interest rate risk is modeled as a CIR process under the Bayesian approach. As a consequence of casting CAT pricing models into our framework, we evaluate the price and expected risk premia of various CAT bond contracts corresponding to clustering of catastrophe risk profiles. Numerical experiments show how these clusters reveal how CAT bond prices and expected risk premia relate to claim frequency and loss severity.

As technology advanced, collecting data via automatic collection devices become popular, thus we commonly face data sets with lengthy variables, especially when these data sets are collected without specific research goals beforehand. It has been pointed out in the literature that the difficulty of high-dimensional classification problems is intrinsically caused by too many noise variables useless for reducing classification error, which offer less benefits for decision-making, and increase complexity, and confusion in model-interpretation. A good variable selection strategy is therefore a must for using such kinds of data well; especially when we expect to use their results for the succeeding applications/studies, where the model-interpretation ability is essential. hus, the conventional classification measures, such as accuracy, sensitivity, precision, cannot be the only performance tasks. In this paper, we propose a probability-based nonparametric multiple-class classification method, and integrate it with the ability of identifying high impact variables for individual class such that we can have more information about its classification rule and the character of each class as well. The proposed method can have its prediction power approximately equal to that of the Bayes rule, and still retains the ability of "model-interpretation." We report the asymptotic properties of the proposed method, and use both synthesized and real data sets to illustrate its properties under different classification situations. We also separately discuss the variable identification, and training sample size determination, and summarize those procedures as algorithms such that users can easily implement them with different computing languages.

This paper deals with the taking into account a given set of realizations as constraints in the Kullback-Leibler minimum principle, which is used as a probabilistic learning algorithm. This permits the effective integration of data into predictive models. We consider the probabilistic learning of a random vector that is made up of either a quantity of interest (unsupervised case) or the couple of the quantity of interest and a control parameter (supervised case). A training set of independent realizations of this random vector is assumed to be given and to be generated with a prior probability measure that is unknown. A target set of realizations of the QoI is available for the two considered cases. The framework is the one of non-Gaussian problems in high dimension. A functional approach is developed on the basis of a weak formulation of the Fourier transform of probability measures (characteristic functions). The construction makes it possible to take into account the target set of realizations of the QoI in the Kullback-Leibler minimum principle. The proposed approach allows for estimating the posterior probability measure of the QoI (unsupervised case) or of the posterior joint probability measure of the QoI with the control parameter (supervised case). The existence and the uniqueness of the posterior probability measure is analyzed for the two cases. The numerical aspects are detailed in order to facilitate the implementation of the proposed method. The presented application in high dimension demonstrates the efficiency and the robustness of the proposed algorithm.

Describing the causal relations governing a system is a fundamental task in many scientific fields, ideally addressed by experimental studies. However, obtaining data under intervention scenarios may not always be feasible, while discovering causal relations from purely observational data is notoriously challenging. In certain settings, such as genomics, we may have data from heterogeneous study conditions, with soft (partial) interventions only pertaining to a subset of the study variables, whose effects and targets are possibly unknown. Combining data from experimental and observational studies offers the opportunity to leverage both domains and improve on the identifiability of causal structures. To this end, we define the interventional BGe score for a mixture of observational and interventional data, where the targets and effects of intervention may be unknown. To demonstrate the approach we compare its performance to other state-of-the-art algorithms, both in simulations and data analysis applications. Prerogative of our method is that it takes a Bayesian perspective leading to a full characterisation of the posterior distribution of the DAG structures. Given a sample of DAGs one can also automatically derive full posterior distributions of the intervention effects. Consequently the method effectively captures the uncertainty both in the structure and the parameter estimates. Codes to reproduce the simulations and analyses are publicly available at github.com/jackkuipers/iBGe

Evaporation calculations are important for the proper management of hydrological resources, such as reservoirs, lakes, and rivers. Data-driven approaches, such as adaptive neuro fuzzy inference, are getting popular in many hydrological fields. This paper investigates the effective implementation of artificial intelligence on the prediction of evaporation for agricultural area. In particular, it presents the adaptive neuro fuzzy inference system (ANFIS) and hybridization of ANFIS with three optimizers, which include the genetic algorithm (GA), firefly algorithm (FFA), and particle swarm optimizer (PSO). Six different measured weather variables are taken for the proposed modelling approach, including the maximum, minimum, and average air temperature, sunshine hours, wind speed, and relative humidity of a given location. Models are separately calibrated with a total of 86 data points over an eight-year period, from 2010 to 2017, at the specified station, located in Arizona, United States of America. Farming lands and humid climates are the reason for choosing this location. Ten statistical indices are calculated to find the best fit model. Comparisons shows that ANFIS and ANFIS–PSO are slightly better than ANFIS–FFA and ANFIS–GA. Though the hybrid ANFIS–PSO (R2= 0.99, VAF = 98.85, RMSE = 9.73, SI = 0.05) is very close to the ANFIS (R2 = 0.99, VAF = 99.04, RMSE = 8.92, SI = 0.05) model, preference can be given to ANFIS, due to its simplicity and easy operation.

Past research on pedestrian trajectory forecasting mainly focused on deterministic predictions which provide only point estimates of future states. These future estimates can help an autonomous vehicle plan its trajectory and avoid collision. However, under dynamic traffic scenarios, planning based on deterministic predictions is not trustworthy. Rather, estimating the uncertainty associated with the predicted states with a certain level of confidence can lead to robust path planning. Hence, the authors propose to quantify this uncertainty during forecasting using stochastic approximation which deterministic approaches fail to capture. The current method is simple and applies Bayesian approximation during inference to standard neural network architectures for estimating uncertainty. The authors compared the predictions between the probabilistic neural network (NN) models with the standard deterministic models. The results indicate that the mean predicted path of probabilistic models was closer to the ground truth when compared with the deterministic prediction. Further, the effect of stochastic dropout of weights and long-term prediction on future state uncertainty has been studied. It was found that the probabilistic models produced better performance metrics like average displacement error (ADE) and final displacement error (FDE). Finally, the study has been extended to multiple datasets providing a comprehensive comparison for each model.

Natural processes may undergo transient periods of nonstationarity which produce lasting change in process behavior across time. When driven by exogeneous influences these changes can be challenging to predict in advance. To circumvent this challenge, works in online (sequential) change detection aim to deduce the occurrence of change in process behavior as it occurs via direct observation of an online data stream. While such changes in process behavior are most commonly modeled via change-points, in which the parameters and/or densities defining an associated process model are assumed to undergo an abrupt and instantaneous transition, changes in the behavior of some processes may occur gradually, taking time to reach their full effect. In such cases change-point models may be ill suited, producing either inaccurate estimates for the timing of these changes or, when this gradual change occurs slowly and when change detection is performed concurrently with model estimation, failing to properly detect change occurrence, as we show empirically in Section 4. This effect can have a significant impact in application, where automated system controls may not be appropriately applied at the correct times, and can result in inaccurate models of process behavior during and after this gradual change.