BayesWatch / deep-kernel-transfer

Official pytorch implementation of the paper "Bayesian Meta-Learning for the Few-Shot Setting via Deep Kernels" (NeurIPS 2020)
https://arxiv.org/abs/1910.05199
197 stars 29 forks source link

whats the role of DKT? How Bayesian is being used in this paper (need under the hood explanation) and how bayesian and DKT are complimenting each other.. #18

Closed Shantan243 closed 1 year ago

Shantan243 commented 1 year ago

Hi, The paper has not explained details of how DKT is creating the difference except in metrics which shows experiement results - Need how DKT is identified as the influencer/factor, what is the role/objective of DKT. Also in papaer, we see Baseline and Baseline++ models , what are these? Could you please give more details/explanations of these Baseline/Baseline++

mpatacchiola commented 1 year ago

Hi @Shantan243

The paper has not explained details of how DKT is creating the difference except in metrics which shows experiement results - Need how DKT is identified as the influencer/factor, what is the role/objective of DKT.

I am not sure what you mean here, can you reformulate the question?

Also in papaer, we see Baseline and Baseline++ models , what are these? Could you please give more details/explanations of these Baseline/Baseline++

Those baselines have been proposed by Chen et al. (2019). They are briefly discussed in the paper (Section 4, Related Work). They use a pretrained backbone that is fine-tuned on a task. Give a look at Chen et al. (2019) for more details.

mpatacchiola commented 1 year ago

Closed for inactivity.