Many key problems in machine learning and data science are routinely modeled as optimization problems and solved via optimization algorithms. With the increase of the volume of data and the size and complexity of the statistical models used to formulate these often ill-conditioned optimization tasks, there is a need for new efficient algorithms able to cope with these challenges. In this thesis, we deal with each of these sources of difficulty in a different way. To efficiently address the big data issue, we develop new methods which in each iteration examine a small random subset of the training data only. To handle the big model issue, we develop methods which in each iteration update a random subset of the model parameters only. Finally, to deal with ill-conditioned problems, we devise methods that incorporate either higher-order information or Nesterov’s acceleration/momentum. In all cases, randomness is viewed as a powerful algorithmic tool that we tune, both in theory and in experiments, to achieve the best results. Our algorithms have their primary application in training supervised machine learning models via regularized empirical risk minimization, which is the dominant paradigm for training such models. However, due to their generality, our methods can be applied in many other fields, including but not limited to data science, engineering, scientific computing, and statistics.

One of the main objectives of evolutionary biology is to understand the reasons behind the maintenance of individual differences in a multitude of traits that influence fitness such as locomotor performance and behaviour. Because locomotor performance sets an "envelope" within which behaviour is expressed, it is likely that a multitude of co-adaptations exists between these two suites of traits. In recent years, a growing number of studies have identified associations of different strength and directions between performance and behaviour. Two main hypotheses have received support, on one hand locomotor performance could be "co-specialized" with behaviour in a manner that behaviour reduces predation risk, such that shyer, less active, less explorative animals should be the best sprinters and the most endurant. On the other hand, locomotor performance could "compensate" for behaviours that lead to increased predation risk, in a way that bolder, more active and explorative animals should be able to sprint faster and for longer. In my thesis I provide a review of published studies that successfully identify associations between locomotor performance and behaviour and classify each association as supporting the co-specialization or compensation hypothesis respectively. I further elaborate on the importance of using repeated measurements and (co)variance partitioning when studying correlations between labile traits. I also discuss one of the main challenges that comes with studying locomotor performance, namely the importance of the variation in motivation, both methodologically, by using different performance tests, but also physiologically, by using blood corticosterone measurements as indicators of such variation.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.