We present the UvA-ILLC submission of the BEER metric to WMT 14 metrics task. BEER is a sentence level metric that can incorporate a large number of features combined in a linear model. Novel contributions are (1) efficient tuning of a large number of features for maximizing correlation with human system ranking, and (2) novel features that give smoother sentence level scores.

The leading policy to conserve forest is protected areas (PAs). Yet, PAs are not a single tool: land users and uses vary by PA type; and public PA strategies vary in the extent of each type and in the determinants of impact for each type, i.e. siting and internal deforestation. Further, across regions and time, strategies respond to pressures (deforestation and political). We estimate deforesta...

We present detailed semi-analytical models for the formation of disk galaxies both in a Universe dominated by dark matter (DM), and in one for which the force law is given by modified Newtonian dynamics (MOND). We tune the models to fit the observed near-infrared Tully-Fisher (TF) relation, and compare numerous predictions of the resulting models with observations. The DM and MOND models are al...