Search
Search Funnelback University
1 -
9 of
9
search results for First graphene-based flexible display produced |u:www.mlmi.eng.cam.ac.uk
where 0
match all words and 9
match some words.
Results that match 4 of 5 words
-
Constrained Bayesian Optimization for Automatic Chemical Design
https://www.mlmi.eng.cam.ac.uk/files/griffiths_dissertation.pdf30 Oct 2019: The first approach is based on optimizing(?) directly and results in a type II maximum likelihood estimate for the hyperparameters [94].This approach can overfit if the data is sparse. ... in conjunction with gradient-based optimization to learn the -
Tradeoffs in Neural Variational Inference
https://www.mlmi.eng.cam.ac.uk/files/cruz_dissertation.pdf30 Oct 2019: 39] provides two versions of this dataset. One consists of theoriginal data they gathered while with the other, the original data is first roughly aligned usinga similarity transformation based on the ... We first croppedthe image to be 178178 before -
Improving Deep Ensembles for Better Deep Uncertainty Quantification
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/improving_deep_ensembles.pdf15 Nov 2021: However, theextrapolations away from training samples, produced by models with these parameter sets candiffer greatly. ... The first dataset examined, chosen for popularity andfast experimentation, is MNIST (LeCun et al., 2010). -
3D Human Motion Synthesis with Recurrent Gaussian Processes
https://www.mlmi.eng.cam.ac.uk/files/mphil_thesis_yeziwei_wang.pdf6 Nov 2019: All above RNN-based models have one common limitation in spite of good results formotion modelling. ... 5] mentions that deep learning structure are mainlyassociated with Restricted Boltzmann Machine (RBM) based models. -
Fairness in Machine Learning withCausal Reasoning Philip Ball…
https://www.mlmi.eng.cam.ac.uk/files/ball_thesis.pdf6 Nov 2019: 4. Applying the CFU metric to a number of observationally based approaches and un-derstanding how these perform counterfactually. ... For example, consider that we wish to grant loans, and make decisions based onwhere an individual lives. -
Understanding Uncertainty in Bayesian Neural Networks
https://www.mlmi.eng.cam.ac.uk/files/mphil_thesis_javier_antoran.pdf18 Nov 2019: We must resort to Monte Carlo (MC) estimatesof the ELBO and gradient based optimisation. ... It was first appliedto BNNs by Hernández-Lobato and Adams (2015). Hernandez-Lobato et al. -
Deep Reinforcement Learning for 3D Molecular Design
https://www.mlmi.eng.cam.ac.uk/files/2021-2022_dissertations/deep_reinforcement_learning_with_3d_molecular_design.pdf25 Nov 2022: First, the policy model constructs a representation of the current canvas thatresembles an enhanced graph. ... 9 Background. approach of MolGym differs from purely graph based molecular representations e.g.,Z. -
Function Constrained Program Synthesis
https://www.mlmi.eng.cam.ac.uk/files/2022_-_2023_dissertations/function_constrained_program_synthesis.pdf17 Nov 2023: function—even when encountering it for the first time—in context, rather than. ... Lever-. aging the success of these architectures, similar transformer-based models were applied. -
Depth Uncertainty Networks for Active Learning
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/depth_uncertainty_networks_for_active_learning_reduced.pdf15 Nov 2021: Finally,we propose a modification to an information-based acquisition strategy, termed truncatedBALD, which mitigates bias caused by model misspecification. ... Bayesian deep learning provides a solution to the first two of these issues.
Search history
Recently clicked results
Recently clicked results
Your click history is empty.
Recent searches
Recent searches
Your search history is empty.