Many transitions scholars underscore the importance of learning in sustainability transitions, but the associated learning processes have hardly been conceptualised. The diverse, well-established research fields related to learning are broadly ignored or loosely applied. In this paper, we systematically explore four interesting learning traditions in terms of their value for gaining an in-depth understanding of learning in sustainability transitions and their relevance for fostering learning, by connecting them to key features of transitions. The selected learning traditions from different disciplinary backgrounds provide valuable insights. None of them sufficiently addresses the complexity of transitions. They include, however, a diversity of relevant learning contexts. We conclude that they have value for investigating new areas such as learning in socio-technological regimes and in later phases of a transition, while enlightening forms of learning that have not yet been fully recognised in transition studies, such as superficial learning, unlearning, and learning to resist change.
MULTIFILE
Dit essay geeft een systeemvisie op het ontwikkelen van embedded software voor slimme systemen: (mobiele) robots en sensornetwerken.
We present a novel architecture for an AI system that allows a priori knowledge to combine with deep learning. In traditional neural networks, all available data is pooled at the input layer. Our alternative neural network is constructed so that partial representations (invariants) are learned in the intermediate layers, which can then be combined with a priori knowledge or with other predictive analyses of the same data. This leads to smaller training datasets due to more efficient learning. In addition, because this architecture allows inclusion of a priori knowledge and interpretable predictive models, the interpretability of the entire system increases while the data can still be used in a black box neural network. Our system makes use of networks of neurons rather than single neurons to enable the representation of approximations (invariants) of the output.
LINK
