Estimating hidden processes from non-linear noisy observations is particularly difficult when the parameters of these processes are not known. This paper adopts a machine learning approach to devise variational Bayesian inference for such scenarios. In particular, a random process generated by the autoregressive moving average (ARMA) linear model is inferred from non-linearity noise observations. The posterior distribution of hidden states are approximated by a set of weighted particles generated by the sequential Monte carlo (SMC) algorithm involving sampling with importance sampling resampling (SISR). Numerical efficiency and estimation accuracy of the proposed inference method are evaluated by computer simulations. Furthermore, the proposed inference method is demonstrated on a practical problem of estimating the missing values in the gene expression time series assuming vector autoregressive (VAR) data model.

Algorithms for Gaussian process, marginal likelihood methods or restricted maximum likelihood methods often require derivatives of log determinant terms. These log determinants are usually parametric with variance parameters of the underlying statistical models. This paper demonstrates that, when the underlying matrix is sparse, how to take the advantage of sparse inversion---selected inversion which share the same sparsity as the original matrix---to accelerate evaluating the derivative of log determinant.

In the previous post we saw what Bayes' Theorem is, and went through an easy, intuitive example of how it works. You can find this post here. If you don't know what Bayes' Theorem is, and you have not had the pleasure to read it yet, I recommend you do, as it will make understanding this present article a lot easier. In this post, we will see the uses of this theorem in Machine Learning. As mentioned in the previous post, Bayes' theorem tells use how to gradually update our knowledge on something as we get more evidence or that about that something.

When explaining probabilistic models, any human-oriented framework for interpretability should take into account how humans understand and interpret probabilities. The psychological and cognitive science communities have long studied this topic (tversky1974judgment), showing, for example, that humans are notoriously bad at incorporating class priors when thinking about probabilities. The classic example of Breast Cancer diagnosis due to eddy1982probabilistic, showed that the majority of subjects (doctors) tended to provide estimates of posterior probabilities roughly one order of magnitude higher that the true values. This phenomenon has been attributed to a neglect of base-rates during reasoning (the base-rate fallacy (bar-hillel1980base)), or instead, to a confusion of inverse conditional probabilities P(A B) and P(B A), one of which needs to be estimated and the other one is provided (the inverse fallacy, (koehler1996base)). Whatever the cause, we argue here that its effect--i.e., that humans often struggle to reason about posterior probabilities--should be taken into account.

Model selection is the problem of choosing one from among a set of candidate models. It is common to choose a model that performs the best on a hold-out test dataset or to estimate model performance using a resampling technique, such as k-fold cross-validation. An alternative approach to model selection involves using probabilistic statistical measures that attempt to quantify both the model performance on the training dataset and the complexity of the model. Examples include the Akaike and Bayesian Information Criterion and the Minimum Description Length. The benefit of these information criterion statistics is that they do not require a hold-out test set, although a limitation is that they do not take the uncertainty of the models into account and may end-up selecting models that are too simple.

A key challenge in reinforcement learning (RL) is how to bala nce exploration and exploitation in order to efficiently learn to make good sequences of decisions in a way that is both computationally tractable and statistically efficient. In the tabular case, the exploration-exploitation problem is well-understood for a number of settings (e.g., finite-hori zon, average reward, infinite horizon with discount), explorati on objectives (e.g., regret minimization and probably approximately correct), and for different algorithmic appro aches, where optimism-under-uncertainty [JOA10, FPLO18] and Thompson sampling (TS) [OBPVR16, Rus19] are the most pop ular principles. For instance, in the finite-horizon setting, [AOM17] and [ZB19] recently derived minimax optim al and structure adaptive regret bounds for optimistic exploration algorithms. TSbased algorithms have mainly b een analyzed in tabular MDPs in terms of Bayesian regret [OBPVR16, OR17, OGNJ17], which assumes that the MDP is s ampled from a known prior distribution. These bounds do not hold against a fixed MDP and algorithms with smal l Bayesian regret may still suffer high regret in some hard-to-learn MDPs within the chosen prior. In the tabu lar setting, frequentist (or worst-case) regret analysis h as been developed for TSbased algorithms both in the average r eward [GM15, AJ17] and finite-horizon case [Rus19]. Despite the fact that TSbased approaches have slightly wor se regret bounds compared to optimism-based algorithms, their empirical performance is often superior [CL11, OR17] . Unfortunately, the performance of tabular exploration met hods rapidly degrades with the number of states and actions, thus making them unfeasible in large or continuous MD Ps.

The framework of cognitively bounded rationality treats problem solving as fundamentally rational, but emphasises that it is constrained by cognitive architecture and the task environment. This paper investigates a simple decision making heuristic, Take The Best (TTB), within that framework. We formulate TTB as a likelihood-based probabilistic model, where the decision strategy arises by probabilistic inference based on the training data and the model constraints. The strengths of the probabilistic formulation, in addition to providing a bounded rational account of the learning of the heuristic, include natural extensibility with additional cognitively plausible constraints and prior information, and the possibility to embed the heuristic as a subpart of a larger probabilistic model. We extend the model to learn cue discrimination thresholds for continuous-valued cues and experiment with using the model to account for biased preference feedback from a boundedly rational agent in a simulated interactive machine learning task.

We consider the fusion of two aerodynamic data sets originating from differing fidelity physical or computer experiments. We specifically address the fusion of: 1) noisy and in-complete fields from wind tunnel measurements and 2) deterministic but biased fields from numerical simulations. These two data sources are fused in order to estimate the \emph{true} field that best matches measured quantities that serves as the ground truth. For example, two sources of pressure fields about an aircraft are fused based on measured forces and moments from a wind-tunnel experiment. A fundamental challenge in this problem is that the true field is unknown and can not be estimated with 100\% certainty. We employ a Bayesian framework to infer the true fields conditioned on measured quantities of interest; essentially we perform a \emph{statistical correction} to the data. The fused data may then be used to construct more accurate surrogate models suitable for early stages of aerospace design. We also introduce an extension of the Proper Orthogonal Decomposition with constraints to solve the same problem. Both methods are demonstrated on fusing the pressure distributions for flow past the RAE2822 airfoil and the Common Research Model wing at transonic conditions. Comparison of both methods reveal that the Bayesian method is more robust when data is scarce while capable of also accounting for uncertainties in the data. Furthermore, given adequate data, the POD based and Bayesian approaches lead to \emph{similar} results.

In this paper we offer a method and algorithm, which make possible fully autonomous (unsupervised) detection of new classes, and learning following a very parsimonious training priming (few labeled data samples only). Moreover, new unknown classes may appear at a later stage and the proposed xClass method and algorithm are able to successfully discover this and learn from the data autonomously. Furthermore, the features (inputs to the classifier) are automatically sub-selected by the algorithm based on the accumulated data density per feature per class. As a result, a highly efficient, lean, human-understandable, autonomously self-learning model (which only needs an extremely parsimonious priming) emerges from the data. To validate our proposal we tested it on two challenging problems, including imbalanced Caltech-101 data set and iRoads dataset. Not only we achieved higher precision, but, more significantly, we only used a single class beforehand, while other methods used all the available classes) and we generated interpretable models with smaller number of features used, through extremely weak and weak supervision.

Bayesian Neural Networks with Latent Variables (BNN+LV's) provide uncertainties in prediction estimates by explicitly modeling model uncertainty (via priors on network weights) and environmental stochasticity (via a latent input noise variable). In this work, we first show that BNN+LV suffers from a serious form of non-identifiability: explanatory power can be transferred between model parameters and input noise while fitting the data equally well. We demonstrate that, as a result, traditional inference methods may yield parameters that reconstruct observed data well but generalize poorly. Next, we develop a novel inference procedure that explicitly mitigates the effects of likelihood non-identifiability during training and yields high quality predictions as well as uncertainty estimates. We demonstrate that our inference method improves upon benchmark methods across a range of synthetic and real datasets.