Learning Personalized Attribute Preference via Multi-Task AUC Optimization

Authors: Zhiyong Yang, Qianqian Xu, Xiaochun Cao, Qingming Huang5660-5667

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical analysis consistently speaks to the efficacy of our proposed method. and Empirical Study Experiment Settings For all the experiments, hyper-parameters are tuned based on the training and validation set(account for 85% of the total instances), and the result on the test set are recorded.
Researcher Affiliation Academia 1SKLOIS, Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China 2School of Cyber Security, University of Chinese Academy of Sciences, Beijing, China 3Key Lab of Intell. Info. Process., Inst. of Comput. Tech., CAS, Beijing, China 4University of Chinese Academy of Sciences, Beijing, China 5Key Laboratory of Big Data Mining and Knowledge Management, CAS, Beijing, China {yangzhiyong, caoxiaochun}@iie.ac.cn, xuqianqian@ict.ac.cn, qmhuang@ucas.ac.cn
Pseudocode No The paper describes its methods but does not contain structured pseudocode or algorithm blocks (clearly labeled algorithm sections or code-like formatted procedures).
Open Source Code Yes The codes are now available 1joshuaas.github.io/publication.html
Open Datasets Yes The Shoes Dataset is collected from (Kovashka and Grauman 2015) which contains 14,658 online shopping images. and The SUN Attributes Dataset (Patterson and Hays 2012), is a well-known large-scale scene attribute dataset with roughly 1,4000 images and a taxonomy of 102 discriminative attributes. Recently, in (Kovashka and Grauman 2015), the personalized annotations over five attributes are collected with hundreds of annotators.
Dataset Splits Yes For all the experiments, hyper-parameters are tuned based on the training and validation set(account for 85% of the total instances), and the result on the test set are recorded.
Hardware Specification No The paper mentions 'our server couldn't finish the program within 1h due to the memory limit (24GB)' but does not provide specific GPU/CPU models, processor types, or detailed computer specifications used for running its experiments.
Software Dependencies No The paper mentions using 'the second last fc layer of the Inception-V3 (Szegedy et al. 2016) network' but does not provide specific software names with version numbers for reproducibility.
Experiment Setup No The paper states 'hyper-parameters are tuned based on the training and validation set' but does not provide specific hyperparameter values, training configurations, or system-level settings in the main text.