Search-based software engineering views software development as a process of searching through the design space for an optimal solution according to some quality criteria. It seems natural to try and build automated implementations of this idea based on concepts from model-driven engineering—using meta-models as characterisations of design spaces and model transformations as algorithms / heuris...

This paper presents a default model-selection procedure for Gaussian graphical models that involves two new developments. First, we develop an objective version of the hyper-inverse Wishart prior for restricted covariance matrices, called the HIW g-prior, and show how it corresponds to the implied fractional prior for covariance selection using fractional Bayes factors. Second, we apply a class...

This paper deals with the detection of multiple changepoints for independent but non identically distributed observations, which are assumed to be modeled by a linear regression with normal errors. The problem has a natural formulation as a model selection problem and the main difficulty for computing model posterior probabilities is that neither the reference priors nor any form of empirical B...

In iterative/incremental software development, software deliverables are built in iterations each iteration providing parts of the required software functionality. To better manage and monitor resources, plan and deliverables, iterations are usually performed during specific time periods, so called “time boxes”. Each time box is further divided into a sequence of stages and a dedicated developm...

For the Multi-Objective Flexible Job Shop Scheduling Problems (MOFJSSP), various optimization models have been designed. In the first part of the research (published in no. 2/2012) we presented the multi-sort production systems, the theoretical aspects of MOFJSSP framework, a specific real JSS process in drugs industry based on over 600 tasks to be optimally scheduled (used as a case study for ...

Sequence-to-sequence neural network models for generation of conversational responses tend to generate safe, commonplace responses (e.g., I don’t know) regardless of the input. We suggest that the traditional objective function, i.e., the likelihood of output (response) given input (message) is unsuited to response generation tasks. Instead we propose using Maximum Mutual Information (MMI) as t...

We propose an objective Bayesian method for the comparison of all Gaussian directed acyclic graphical models defined on a given set of variables. The method, which is based on the notion of fractional Bayes factor, requires a single default (typically improper) prior on the space of unconstrained covariance matrices, together with a prior sample size hyper-parameter, which can be set to its min...

In unsupervised data generation tasks, besides the generation of a sample based on previous observations, one would often like to give hints to the model in order to bias the generation towards desirable metrics. We propose a method that combines Generative Adversarial Networks (GANs) and reinforcement learning (RL) in order to accomplish exactly that. While RL biases the data generation proces...

We discuss higher-order approximations to the marginal posterior distribution for a scalar parameter of interest in the presence of nuisance parameters. These higher-order approximations are obtained using a suitable matching prior. The proposed procedure has several advantages since it does not require the elicitation on the nuisance parameter, neither numerical integration or MCMC simulation,...

The tuning of parameters in climate models is essential to provide reliable long-term forecasts of Earth system behaviour. We apply a multi-objective optimization algorithm to the problem of parameter estimation in climate models. This optimization process involves the iterative evaluation of response surface models (RSMs), followed by the execution of multiple Earth system simulations. These c...