Bayesian optimization (BO) is a widely-used method for optimizing expensive (to evaluate) problems. At the core of most BO methods is the modeling of the objective function using a Gaussian Process (GP) whose covariance is selected from a set of standard covariance functions. From a weight-space view, this models the objective as a linear function in a feature space implied by the given covariance K, with an arbitrary Gaussian weight prior w ~ormdist (0,I). In many practical applications there is data available that has a similar (covariance) structure to the objective, but which, having different form, cannot be used directly in standard transfer learning. In this paper we show how such auxiliary data may be used to construct a GP covariance corresponding to a more appropriate weight prior for the objective function. Building on this, we show that we may accelerate BO by modeling the objective function using this (learned) weight prior, which we demonstrate on both test functions and a practical application to short-polymer fibre manufacture.

We present a new optical sample of three Supernova Remnants (SNRs) and 16 Supernova Remnant (SNR) candidates in the Large Magellanic Cloud (LMC). These objects were originally selected using deep H α, [S ii], and [O iii] narrow-band imaging. Most of the newly found objects are located in less dense regions, near or around the edges of the LMC’s main body. Together with previously suggested MCSNR J0541–6659, we confirm the SNR nature for two additional new objects: MCSNR J0522–6740 and MCSNR J0542–7104. Spectroscopic follow-up observations for 12 of the LMC objects confirm high [S ii]/H α emission-line ratios ranging from 0.5 to 1.1. We consider the candidate J0509–6402 to be a special example of the remnant of a possible type Ia Supernova (SN) which is situated some 2° (∼1.75 kpc) north from the main body of the LMC. We also find that the SNR candidates in our sample are significantly larger in size than the currently known LMC SNRs by a factor of ∼2. This could potentially imply that we are discovering a previously unknown but predicted, older class of large LMC SNRs that are only visible optically. Finally, we suggest that most of these LMC SNRs are residing in a very rarefied environment towards the end of their evolutionary span where they become less visible to radio and X-ray telescopes.