Nonparametric Stochastic Contextual Bandits
Authors: Melody Guan, Heinrich Jiang
AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Finally, we experimentally show the improvement of our algorithm over existing approaches for both simulated tasks and MNIST image classification. |
| Researcher Affiliation | Collaboration | Melody Y. Guan Stanford University 450 Serra Mall Stanford, California 94305 mguan@stanford.edu Heinrich Jiang Google 1600 Amphitheatre Pwky Mountain View, California 94043 heinrich.jiang@gmail.com |
| Pseudocode | Yes | Algorithm 1 Uniform Sampling 1: Parameters: T, total number of time steps. 2: For each arm i of the K arms: 3: For each time step t [ (i 1)T K ]: 4: Pull arm It := i. 5: Define fi : X R to be the k-NN regression estimator from the sampled context and reward observations for each i [K]. |
| Open Source Code | No | The paper does not provide any concrete access information (e.g., repository link, explicit statement of release, or mention of code in supplementary materials) for the source code of the described methodology. |
| Open Datasets | Yes | Image Classification Experiments We extend our experiments to image classification of the canonical MNIST dataset, which consists of 60k training images and 10k test images of isolated, normalized, handwritten digits. |
| Dataset Splits | Yes | We tuned other hyperparameters using grid search on a validation set of size 1k using grid search and we evaluate performance of our models on a test set of size 1k. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., CPU/GPU models, memory) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers (e.g., programming language versions, library versions) needed to replicate the experiment. |
| Experiment Setup | Yes | We found that a confidence level of 0.1 worked well for all settings. |