Fine-Grained Machine Teaching with Attention Modeling
Authors: Jiacheng Liu, Xiaofeng Hou, Feilong Tang2585-2592
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | For corroborating our theoretical findings, we conduct extensive experiments with both synthetic datasets and real datasets. Our experimental results verify the effectiveness of AMT algorithms. |
| Researcher Affiliation | Academia | Jiacheng Liu, Xiaofeng Hou, Feilong Tang Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, China {liujiacheng, xfhelen}@sjtu.edu.cn, tang-fl@cs.sjtu.edu.cn |
| Pseudocode | Yes | Algorithm 1 Attentive Machine Teaching Algorithm (AMT) ... Algorithm 2 Blackbox AMT Algorithm |
| Open Source Code | No | The paper does not provide any explicit statement or link for the open-sourcing of their code. |
| Open Datasets | Yes | The first is 2D Gaussian dataset, which is a synthetic two-dimensional dataset drawn from Gaussian distribution for selected sample visualization. The second is 10D and 100D Gaussian dataset, which is a synthetic dataset for validating the effectiveness of the proposed method in medium and high dimensions. The third is the hate speech detection dataset (Davidson et al. 2017), which contains three categories including hate speech , offensive language and neither . |
| Dataset Splits | No | The paper mentions 'testing dataset' and 'training data points' but does not explicitly describe train/validation/test splits, sample counts for each split, or cross-validation setup details required for reproduction. For instance, it says 'We use 20% of the data as the testing dataset' for hate speech but doesn't specify how the remaining 80% is split between training and validation, nor does it specify splits for Gaussian datasets. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., CPU, GPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions using a 'two-layer GCN network' and 'softmax', 'ReLU' functions but does not specify any software names with version numbers (e.g., TensorFlow 2.x, PyTorch 1.x, Python 3.x, CUDA versions). |
| Experiment Setup | Yes | The learning rate is set to 0.004 and the initial concept w0 for all the schemes is set with the standard normal distribution. |