Positive and Unlabeled Learning via Loss Decomposition and Centroid Estimation
Authors: Hong Shi, Shaojun Pan, Jian Yang, Chen Gong
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We intensively validate our approach on synthetic dataset, UCI benchmark datasets and real-world datasets, and the experimental results firmly demonstrate the effectiveness of our approach when compared with other state-of-the-art PU learning methodologies. In this section, we perform exhaustive experiments on one synthetic dataset, seven publicly available benchmark datasets and two real-world datasets. |
| Researcher Affiliation | Academia | Hong Shi, Shaojun Pan, Jian Yang and Chen Gong School of Computer Science and Engineering, Nanjing University of Science and Technology Jiangsu Key Laboratory of Image and Video Understanding for Social Security chen.gong@njust.edu.cn |
| Pseudocode | Yes | Algorithm 1 Median-of-means estimator of corrupted negative mean; Algorithm 2 Loss Decomposition and Centroid Estimation (LDCE) algorithm for PU learning. |
| Open Source Code | No | The paper does not provide any information about the availability of the source code for the proposed methodology. |
| Open Datasets | Yes | We intensively validate our approach on synthetic dataset, UCI benchmark datasets and real-world datasets... We also conduct the experiments on two real-world datasets... The USPS 1 dataset was adopted... 1http://www.cad.zju.edu.cn/home/dengcai/Data/MLData.html... Hockey Fight 2 dataset... 2http://visilab.etsii.uclm.es/personas/oscar/Fight Detection/index.html... seven datasets from UCI machine learning repository [Merz and Murphy, 1998]. |
| Dataset Splits | Yes | For each of the dataset illustrated in Table 1, we randomly pick 80% of the data for training and the rest 20% examples are used for testing. In our experiment, we conduct 5-fold cross validation on all comparators and their mean test accuracies over the five trials are reported in Table 1. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used to run the experiments. |
| Software Dependencies | No | The paper does not list specific software dependencies with version numbers, such as programming languages, libraries, or frameworks used for implementation or experimentation. |
| Experiment Setup | No | The paper mentions a regularization term λ and an approximation parameter β, and that 'gradient descent method' is used. However, it does not provide specific values for these parameters (e.g., learning rate, batch size for gradient descent, specific value for λ, or how β is determined beyond 'cross-validation'), initial weights, or other detailed training settings required for reproduction. |