Bayesian Model-Agnostic Meta-Learning

Authors: Jaesik Yoon, Taesup Kim, Ousmane Dia, Sungwoong Kim, Yoshua Bengio, Sungjin Ahn

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiment results show the accuracy and robustness of the proposed method in sinusoidal regression, image classification, active learning, and reinforcement learning. We evaluated our proposed model (BMAML) in various few-shot learning tasks: sinusoidal regression, image classification, active learning, and reinforcement learning.
Researcher Affiliation Collaboration 1Element AI, 2MILA Université de Montréal, 3SAP, 4Kakao Brain, 5CIFAR Senior Fellow, 6Rutgers University
Pseudocode Yes Algorithm 1 MAML, Algorithm 2 Bayesian Fast Adaptation, Algorithm 3 Bayesian Meta-Learning with Chaser Loss (BMAML)
Open Source Code No The paper does not include any explicit statements about releasing source code for the described methodology, nor does it provide a link to a code repository.
Open Datasets Yes We test and compare the models on the same Mu Jo Co continuous control tasks (Todorov et al., 2012) as are used in Finn et al. (2017). mini Imagenet classification task (Vinyals et al., 2016).
Dataset Splits Yes The whole dataset of tasks is divided into training/validation/test tasksets, and the dataset of each task is further divided into task-training/task-validation/task-test datasets. The entire classes are split into 64, 12, and 24 classes for meta-train, meta-validation, and meta-test, respectively.
Hardware Specification No The paper does not provide specific hardware details such as exact GPU/CPU models, processor types, or memory amounts used for running its experiments. It mentions general terms like 'neural network' but no hardware specifications.
Software Dependencies No The paper mentions methods and algorithms like 'RBF kernel', 'Adam', 'TRPO', and 'REINFORCE', but does not provide specific software library names with version numbers (e.g., Python, PyTorch, TensorFlow versions) that would be needed to replicate the experiment.
Experiment Setup Yes For the regression model, we used a neural network with 3 layers each of which consists of 40 hidden units. We split the network architecture into the feature extractor layers and the classifier. The feature extractor is a convolutional network with 5 hidden layers with 64 filters. The classifier is a single-layer fully-connected network with softmax output. In our experiments, setting n and s to a small number like n = s = 1 worked well.