Search
Search Funnelback University
- Refined by:
- Date: Past fortnight
1 -
10 of
12
search results for KaKaoTalk:ZA31 24 24 |u:www.mlmi.eng.cam.ac.uk
where 0
match all words and 12
match some words.
Results that match 2 of 3 words
-
Efficiently-Parametrised Approximate Posteriors in Pseudo-Point…
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/efficiently_parametrized_approximate_posteriors.pdf15 Nov 2021: exp(L1)p(u)q(u)q(u). du. (L1 log p(u)log q(u))q(u)du. = Eq(u) (L1 log p(u)log q(u)) L2 (2.24). ... L2 in Hensman et al. (2013) as follows (see equation 2.24 and 2.20 for the definition of L2and L1 respectively):. -
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: decaying priors. 604.24 Overfitting bias for DUNs with uniform and decaying priors trained with R̃. ... w = t(φ,ε) = µ σ ε ; ε N(0,I), (2.24). -
GPT-3 for Few-Shot Dialogue State Tracking
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/gpt_3_for_few_shot_dialogue_state_tracking.pdf15 Nov 2021: 23. 3.4 Experimental Setup. 243.4.1 Model Configuration. 243.4.2 Experiment Size. 24. -
Mitigating Gender Bias in Dialogue Generation Gabrielle (Ming Yi) ...
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/mitigating_gender_bias_in_dialogue_generation.pdf15 Nov 2021: 23. 3.6 Token counts of GB-Ctrl validation data. 24. 3.7 Toxicity in GB-Ctrl finetuning data. ... 24. 3.8 Size of GBS-Ctrl finetuning datasets. 26. 3.9 StereoSet examples. -
Flow Field and Shape Inference in Magnetic Resonance Velocimetry…
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/flow_field_and_shape_inference_reduced.pdf15 Nov 2021: 24. 3.3 Flow at 100 Reynolds Number. 31. 3.3.1 Flow at 100 Reynolds Number: Assumptions. ... 41. 3.25 Absolute discrepancy between the reconstruction and ground truth z-velocityfields from Figure 3.24. -
Conditional Neural Processes and Semi-Supervised Learning
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/conditional_neural_processes_and_semi-supervised_learning_0.pdf15 Nov 2021: Conditional Neural Processes andSemi-Supervised Learning. Max-Olivier Van Bastelaer. Department of EngineeringUniversity of Cambridge. This dissertation is submitted for the degree ofMaster of Philosophy in Machine Learning and Machine Intelligence. -
A model-based design tool for 3D GUI layout design that accommodates…
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/a_model-based_design_tool.pdf15 Nov 2021: This ideais often traced back to the study by Hess and Polt [24] demonstrating correlation betweenpupil size and mental activity in the form of simple multiplication problems. -
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: Improving Deep Ensembles for BetterDeep Uncertainty Quantification. Ginte Petrulionyte. Department of EngineeringUniversity of Cambridge. This dissertation is submitted for the degree ofMaster of Philosophy in Machine Learning and Machine -
Building a Conversational User Simulator Using Generative Adversarial …
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/building_a_conversational_user_simulator.pdf15 Nov 2021: 244.3.1 Policy Training. 244.3.2 Policy Evaluation. 24. 5 Adversarial Training Experiments 265.1 MLE Pre-Training. -
Causal Representation Learning for Latent Space Optimization
https://www.mlmi.eng.cam.ac.uk/files/2020-2021_dissertations/causal_representation_learning_for_latent_space_optimization.pdf15 Nov 2021: Causal Representation Learning forLatent Space Optimization. Wenlin Chen. Department of EngineeringUniversity of Cambridge. This dissertation is submitted for the degree ofMaster of Philosophy in Machine Learning and Machine Intelligence. St
Refine your results
clear all
Date
Search history
Recently clicked results
Recently clicked results
Your click history is empty.
Recent searches
Recent searches
Your search history is empty.