Search

This paper proposes Bayesian Adaptive Trials (BAT) as both an efficient method to conduct trials and a unifying framework for evaluation social policy interventions, addressing limitations inherent in traditional methods such as Randomized Controlled Trials (RCT). Recognizing the crucial need for evidence-based approaches in public policy, the proposal aims to lower barriers to the adoption of evidence-based methods and align evaluation processes more closely with the dynamic nature of policy cycles. BATs, grounded in decision theory, offer a dynamic, ``learning as we go'' approach, enabling the integration of diverse information types and facilitating a continuous, iterative process of policy evaluation. BATs' adaptive nature is particularly advantageous in policy settings, allowing for more timely and context-sensitive decisions. Moreover, BATs' ability to value potential future information sources positions it as an optimal strategy for sequential data acquisition during policy implementation. While acknowledging the assumptions and models intrinsic to BATs, such as prior distributions and likelihood functions, the paper argues that these are advantageous for decision-makers in social policy, effectively merging the best features of various methodologies.

Open-World Compositional Zero-Shot Learning (OW-CZSL) aims to recognize new compositions of seen attributes and objects. In OW-CZSL, methods built on the conventional closed-world setting degrade severely due to the unconstrained OW test space. While previous works alleviate the issue by pruning compositions according to external knowledge or correlations in seen pairs, they introduce biases that harm the generalization. Some methods thus predict state and object with independently constructed and trained classifiers, ignoring that attributes are highly context-dependent and visually entangled with objects. In this paper, we propose a novel Distilled Reverse Attention Network to address the challenges. We also model attributes and objects separately but with different motivations, capturing contextuality and locality, respectively. We further design a reverse-and-distill strategy that learns disentangled representations of elementary components in training data supervised by reverse attention and knowledge distillation. We conduct experiments on three datasets and consistently achieve state-of-the-art (SOTA) performance.

Recent advances in recommender systems have proved the potential of Reinforcement Learning (RL) to handle the dynamic evolution processes between users and recommender systems. However, learning to train an optimal RL agent is generally impractical with commonly sparse user feedback data in the context of recommender systems. To circumvent the lack of interaction of current RL-based recommender systems, we propose to learn a general Model-Agnostic Counterfactual Synthesis (MACS) Policy for counterfactual user interaction data augmentation. The counterfactual synthesis policy aims to synthesise counterfactual states while preserving significant information in the original state relevant to the user's interests, building upon two different training approaches we designed: learning with expert demonstrations and joint training. As a result, the synthesis of each counterfactual data is based on the current recommendation agent's interaction with the environment to adapt to users' dynamic interests. We integrate the proposed policy Deep Deterministic Policy Gradient (DDPG), Soft Actor Critic (SAC) and Twin Delayed DDPG in an adaptive pipeline with a recommendation agent that can generate counterfactual data to improve the performance of recommendation. The empirical results on both online simulation and offline datasets demonstrate the effectiveness and generalisation of our counterfactual synthesis policy and verify that it improves the performance of RL recommendation agents., Comment: Accepted by IEEE TNNLS

Forecasting models have been influential in shaping decision-making in the COVID-19 pandemic. However, there is concern that their predictions may have been misleading. Here, we dissect the predictions made by four models for the daily COVID-19 death counts between March 25 and June 5 in New York state, as well as the predictions of ICU bed utilisation made by the influential IHME model. We evaluated the accuracy of the point estimates and the accuracy of the uncertainty estimates of the model predictions. First, we compared the "ground truth" data sources on daily deaths against which these models were trained. Three different data sources were used by these models, and these had substantial differences in recorded daily death counts. Two additional data sources that we examined also provided different death counts per day. For accuracy of prediction, all models fared very poorly. Only 10.2% of the predictions fell within 10% of their training ground truth, irrespective of distance into the future. For accurate assessment of uncertainty, only one model matched relatively well the nominal 95% coverage, but that model did not start predictions until April 16, thus had no impact on early, major decisions. For ICU bed utilisation, the IHME model was highly inaccurate; the point estimates only started to match ground truth after the pandemic wave had started to wane. We conclude that trustworthy models require trustworthy input data to be trained upon. Moreover, models need to be subjected to prespecified real time performance tests, before their results are provided to policy makers and public health officials.

This paper provides a formal evaluation of the predictive performance of a model (and its various updates) developed by the Institute for Health Metrics and Evaluation (IHME) for predicting daily deaths attributed to COVID19 for each state in the United States. The IHME models have received extensive attention in social and mass media, and have influenced policy makers at the highest levels of the United States government. For effective policy making the accurate assessment of uncertainty, as well as accurate point predictions, are necessary because the risks inherent in a decision must be taken into account, especially in the present setting of a novel disease affecting millions of lives. To assess the accuracy of the IHME models, we examine both forecast accuracy as well as the predictive performance of the 95% prediction intervals provided by the IHME models. We find that the initial IHME model underestimates the uncertainty surrounding the number of daily deaths substantially. Specifically, the true number of next day deaths fell outside the IHME prediction intervals as much as 70% of the time, in comparison to the expected value of 5%. In addition, we note that the performance of the initial model does not improve with shorter forecast horizons. Regarding the updated models, our analyses indicate that the later models do not show any improvement in the accuracy of the point estimate predictions. In fact, there is some evidence that this accuracy has actually decreased over the initial models. Moreover, when considering the updated models, while we observe a larger percentage of states having actual values lying inside the 95% prediction intervals (PI), our analysis suggests that this observation may be attributed to the widening of the PIs. The width of these intervals calls into question the usefulness of the predictions to drive policy making and resource allocation.

We present a method for the joint analysis of a panel of possibly nonstationary time series. The approach is Bayesian and uses a covariate-dependent infinite mixture model to incorporate multiple time series, with mixture components parameterized by a time varying mean and log spectrum. The mixture components are based on AdaptSPEC, a nonparametric model which adaptively divides the time series into an unknown number of segments and estimates the local log spectra by smoothing splines. We extend AdaptSPEC to handle missing values, a common feature of time series which can cause difficulties for nonparametric spectral methods. A second extension is to allow for a time varying mean. Covariates, assumed to be time-independent, are incorporated via the mixture weights using the logistic stick breaking process. The model can estimate time varying means and spectra at observed and unobserved covariate values, allowing for predictive inference. Estimation is performed by Markov chain Monte Carlo (MCMC) methods, combining data augmentation, reversible jump, and Riemann manifold Hamiltonian Monte Carlo techniques. We evaluate the methodology using simulated data, and describe applications to Australian rainfall data and measles incidence in the US. Software implementing the method proposed in this paper is available in the R package BayesSpec.

We propose a scalable framework for inference in an inhomogeneous Poisson process modeled by a continuous sigmoidal Cox process that assumes the corresponding intensity function is given by a Gaussian process (GP) prior transformed with a scaled logistic sigmoid function. We present a tractable representation of the likelihood through augmentation with a superposition of Poisson processes. This view enables a structured variational approximation capturing dependencies across variables in the model. Our framework avoids discretization of the domain, does not require accurate numerical integration over the input space and is not limited to GPs with squared exponential kernels. We evaluate our approach on synthetic and real-world data showing that its benefits are particularly pronounced on multivariate input settings where it overcomes the limitations of mean-field methods and sampling schemes. We provide the state of-the-art in terms of speed, accuracy and uncertainty quantification trade-offs.

Discrimination between non-stationarity and long-range dependency is a difficult and long-standing issue in modelling financial time series. This paper uses an adaptive spectral technique which jointly models the non-stationarity and dependency of financial time series in a non-parametric fashion assuming that the time series consists of a finite, but unknown number, of locally stationary processes, the locations of which are also unknown. The model allows a non-parametric estimate of the dependency structure by modelling the auto-covariance function in the spectral domain. All our estimates are made within a Bayesian framework where we use aReversible Jump Markov Chain Monte Carlo algorithm for inference. We study the frequentist properties of our estimates via a simulation study, and present a novel way of generating time series data from a nonparametric spectrum. Results indicate that our techniques perform well across a range of data generating processes. We apply our method to a number of real examples and our results indicate that several financial time series exhibit both long-range dependency and non-stationarity., Comment: 8 pages

The rigorous quantification of uncertainty in geophysical inversions is a challenging problem. Inversions are often ill-posed and the likelihood surface may be multi-modal; properties of any single mode become inadequate uncertainty measures, and sampling methods become inefficient for irregular posteriors or high-dimensional parameter spaces. We explore the influences of different choices made by the practitioner on the efficiency and accuracy of Bayesian geophysical inversion methods that rely on Markov chain Monte Carlo sampling to assess uncertainty, using a multi-sensor inversion of the three-dimensional structure and composition of a region in the Cooper Basin of South Australia as a case study. The inversion is performed using an updated version of the Obsidian distributed inversion software. We find that the posterior for this inversion has complex local covariance structure, hindering the efficiency of adaptive sampling methods that adjust the proposal based on the chain history. Within the context of a parallel-tempered Markov chain Monte Carlo scheme for exploring high-dimensional multi-modal posteriors, a preconditioned Crank-Nicholson proposal outperforms more conventional forms of random walk. Aspects of the problem setup, such as priors on petrophysics or on 3-D geological structure, affect the shape and separation of posterior modes, influencing sampling performance as well as the inversion results. Use of uninformative priors on sensor noise can improve inversion results by enabling optimal weighting among multiple sensors even if noise levels are uncertain. Efficiency could be further increased by using posterior gradient information within proposals, which Obsidian does not currently support, but which could be emulated using posterior surrogates., Comment: 17 pages, 5 figures, 1 table; submitted to Geoscientific Model Development

Bayesian neural learning feature a rigorous approach to estimation and uncertainty quantification via the posterior distribution of weights that represent knowledge of the neural network. This not only provides point estimates of optimal set of weights but also the ability to quantify uncertainty in decision making using the posterior distribution. Markov chain Monte Carlo (MCMC) techniques are typically used to obtain sample-based estimates of the posterior distribution. However, these techniques face challenges in convergence and scalability, particularly in settings with large datasets and network architectures. This paper address these challenges in two ways. First, parallel tempering is used used to explore multiple modes of the posterior distribution and implemented in multi-core computing architecture. Second, we make within-chain sampling schemes more efficient by using Langevin gradient information in forming Metropolis-Hastings proposal distributions. We demonstrate the techniques using time series prediction and pattern classification applications. The results show that the method not only improves the computational time, but provides better prediction or decision making capabilities when compared to related methods., Comment: In review. Software: https://github.com/sydney-machine-learning/parallel-tempering-neural-net

Estimating the impact of environmental processes on vertical reef development in geological time is a very challenging task. pyReef-Core is a deterministic carbonate stratigraphic forward model designed to simulate the key biological and environmental processes that determine vertical reef accretion and assemblage changes in fossil reef drill cores. We present a Bayesian framework called Bayesreef for the estimation and uncertainty quantification of parameters in pyReef-Core that represent environmental conditions affecting the growth of coral assemblages on geological timescales. We demonstrate the existence of multimodal posterior distributions and investigate the challenges of sampling using Markov chain Monte-Carlo (MCMC) methods, which includes parallel tempering MCMC. We use synthetic reef-core to investigate fundamental issues and then apply the methodology to a selected reef-core from the Great Barrier Reef in Australia. The results show that Bayesreef accurately estimates and provides uncertainty quantification of the selected parameters that represent the environment and ecological conditions in pyReef-Core. Bayesreef provides insights into the complex posterior distributions of parameters in pyReef-Core, which provides the groundwork for future research in this area.

The Bayesian paradigm is becoming an increasingly popular framework for estimation and uncertainty quantification of unknown parameters in geo-physical inversion problems. Badlands is a basin and landscape evolution forward model for simulating topography evolution at a large range of spatial and time scales. Our previous work presented Bayeslands that used the Bayesian paradigm to make inference for unknown parameters in the Badlands model using Markov chain Monte Carlo (MCMC) sampling. Bayeslands faced challenges in convergence due to multi-modal posterior distributions in the selected parameters of Badlands. Parallel tempering is an advanced MCMC method suited for irregular and multi-modal posterior distributions. In this paper, we extend Bayeslands using parallel tempering (PT-Bayeslands) with high performance computing to address previous limitations in parameter space exploration in the context of the computationally expensive Badlands model. Our results show that PT-Bayeslands not only reduces the computation time, but also provides an improvement of the sampling for multi-modal posterior distributions. This provides an improvement over Bayeslands which used single chain MCMC that face difficulties in convergence and can lead to misleading inference. This motivates its usage in large-scale basin and landscape evolution models.

Bayesian inference provides a rigorous methodology for estimation and uncertainty quantification of parameters in geophysical forward models. Badlands (basin and landscape dynamics model) is a landscape evolution model that simulates topography development at various space and time scales. Badlands consists of a number of geophysical parameters that needs estimation with appropriate uncertainty quantification; given the observed present-day ground truth such as surface topography and the stratigraphy of sediment deposition through time. The inference of unknown parameters is challenging due to the scarcity of data, sensitivity of the parameter setting and complexity of the Badlands model. In this paper, we take a Bayesian approach to provide inference using Markov chain Monte Carlo sampling (MCMC). We present \textit{Bayeslands}; a Bayesian framework for Badlands that fuses information obtained from complex forward models with observational data and prior knowledge. As a proof-of-concept, we consider a synthetic and real-world topography with two parameters for Bayeslands inference, namely precipitation and erodibility. The results of the experiments show that Bayeslands yields a promising distribution of the parameters. Moreover, we demonstrate the challenge in sampling irregular and multi-modal posterior distributions using a likelihood surface that has a range of sub-optimal modes.

Naveau et al. (2016) have recently developed a class of methods, based on extreme-value theory (EVT), for capturing low, moderate, and heavy rainfall simultaneously, without the need to choose a threshold typical to EVT methods. We analyse the performance of Naveau et al.'s methods, along with mixtures of gamma distributions, by fitting them to marginal non-zero rainfall from 16,968 sites spanning the Australian continent and which represent a wide variety of rainfall patterns. Performance is assessed by the distribution across sites of the log ratios of each method's estimated quantiles and the empirical quantiles. We do so for quantiles corresponding to low, moderate, and heavy rainfall. Under this metric, mixtures of three and four gamma distributions outperform Naveau et al's methods for small and moderate rainfall, and provide equivalent fits for heavy rainfall.

The increasing ease of data capture and storage has led to a corresponding increase in the choice of data, the type of analysis performed on that data, and the complexity of the analysis performed. The main contribution of this paper is to show that the subjective choice of data and analysis methodology substantially impacts the identification of factors and outcomes of observational studies. This subjective variability of inference is at the heart of recent discussions around irreproducibility in scientific research. To demonstrate this subjective variability, data is taken from an existing study, where interest centres on understanding the factors associated with a young adult's propensity to fall into the category of `not in employment, education or training' (NEET). A fully probabilistic analysis is performed, set in a Bayesian framework and implemented using Reversible Jump Markov chain Monte Carlo (RJMCMC). The results show that different techniques lead to different inference but that models consisting of different factors often have the same predictive performance, whether the analysis is frequentist or Bayesian, making inference problematic. We demonstrate how the use of prior distributions in Bayesian techniques can be used to as a tool for assessing a factor's importance., Comment: 14 pages, 5 figures

This article presents a Bayesian approach for predicting and identifying the factors which most influence an individual's propensity to fall into the category of Not in Employment Education or Training (NEET). The approach partitions the covariates into two groups: those which have the potential to be changed as a result of an intervention strategy and those which must be controlled for. This partition allows us to develop models and identify important factors conditional on the control covariates, which is useful for clinicians and policy makers who wish to identify potential intervention strategies. Using the data obtained by O'Dea (2014) we compare the results from this approach with the results from O'Dea (2014) and with the results obtained using the Bayesian variable selection procedure of Lamnisos (2009) when the covariates are not partitioned. We find that the relative importance of predictive factors varies greatly depending upon the control covariates. This has enormous implications when deciding on what interventions are most useful to prevent young people from being NEET., Comment: 5 pages, 2 figures, 2 tables, NIPS 2016 Workshop on Machine Learning for Health