Iterative Metric Learning for Imbalance Data Classification
Authors: Nan Wang, Xibin Zhao, Yu Jiang, Yue Gao
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | To evaluate the effectiveness of the proposed method, we have conducted experiments on two groups of dataset, i.e., the NASA Metrics Data Program (NASA) dataset and UCI Machine Learning Repository (UCI) dataset. Experimental results and comparisons with state-of-the-art methods have exhibited better performance of our proposed method. |
| Researcher Affiliation | Academia | BNRist, KLISS, School of Software, Tsinghua University, China |
| Pseudocode | No | The paper describes the method verbally and provides figures to illustrate the framework and processes, but it does not include any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for the described methodology is publicly available. |
| Open Datasets | Yes | In our experiments, we employ the widely used seven data from NASA Metrics Data Program (NASA) dataset [Menzies et al., 2007], including CM1, KC3, MC2, MW1, PC1, PC3, PC4 and nine data from binary UCI Machine Learning Repository (UCI)[Lichman, 2013], including australian, haberman, heartstatlog, ionosphere, Liver Disorders, sonar, SPET, SPECTF, wdbc to evaluate the performance of our method. |
| Dataset Splits | No | The paper states that datasets are randomly divided into "training datasets and testing datasets", but it does not explicitly mention a separate "validation" dataset split or provide details for such a split. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware (e.g., GPU models, CPU types, memory) used to run the experiments. |
| Software Dependencies | No | The paper mentions using KNN as a classifier but does not specify any software dependencies or their version numbers (e.g., Python, TensorFlow, PyTorch versions). |
| Experiment Setup | Yes | Moreover, we choose the k-nearest neighbors (KNN) algorithm as classifier, and compare the performance of our method with KNN using data space which is handled by different times of iterations. Considering the sparsity of data space, it is difficult to determine the matching rate for training samples matching process. In order to select the most effective matching ratio, we vary K from 0.5 to 0.9, and set matching ratio as 0.8 by experimental results. |