Multi-Instance Learning with Key Instance Shift

Authors: Ya-Lin Zhang, Zhi-Hua Zhou

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The experiments are performed on both synthetic data and real world datasets. We compare the proposed method with many state-of-the-art algorithms...
Researcher Affiliation Academia Ya-Lin Zhang and Zhi-Hua Zhou National Key Laboratory for Novel Software Technology, Nanjing University Collaborative Innovation Center of Novel Software Technology and Industrialization Nanjing 210023, China {zhangyl, zhouzh}@lamda.nju.edu.cn
Pseudocode Yes Algorithm 1 Instance Prototype Learning for MIKI
Open Source Code No The paper does not provide a direct statement or link for the open-source code of the proposed method.
Open Datasets No The paper mentions using '20 Newsgroups corpora' and 'benchmark datasets' (Musk1, Musk2, Elephant, Fox, Tiger) but does not provide specific links, DOIs, or formal citations for public access to the exact datasets or splits used.
Dataset Splits Yes We follow the typically deliberately biased sampling procedure [Zadrozny, 2004] to separate the bags into disjoint training and test sets. Specifically, we define a random variable si for each bag... Pr(si = 1|xi SC1) = a , Pr(si = 1|xi SC2) = b . Here, SC1 and SC2 denote different sub-concepts (i.e.,P1 and P2) and a = 0.8, b = 0.2. Furthermore, we conduct ten times 10-fold cross validations as previous studies done.
Hardware Specification No The paper does not provide specific hardware details (like CPU/GPU models or memory) used for running the experiments.
Software Dependencies No The paper mentions 'RBF kernel is used for all SVM-based methods' and 'multi-class SVM model' but does not specify any software names with version numbers (e.g., Python, scikit-learn, PyTorch versions) used for implementation.
Experiment Setup Yes For MIKI, we simply set K to 5 for synthetic dataset and 10 for the other datasets without any tuning, and set max iteration to 5 to accelerate the method. Other parameters are selected via 5-fold cross validation.