Minimax Optimal Alternating Minimization for Kernel Nonparametric Tensor Learning

Authors: Taiji Suzuki, Heishiro Kanagawa, Hayato Kobayashi, Nobuyuki Shimizu, Yukihiro Tagami

NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 7 Numerical experiments We numerically compare the following methods in multitask learning problems (Eq. (2)): ... 7.1 Restaurant data ... Fig. 1 shows the relative MSE ... Fig. 2 shows the performance comparison ... 7.2 Online shopping data ... The result is shown in Fig. 3, which presents the validation error (MSE) against the size of the training data.
Researcher Affiliation Collaboration Taiji Suzuki , Heishiro Kanagawa , Department of Mathematical and Computing Science, Tokyo Institute of Technology PRESTO, Japan Science and Technology Agency Center for Advanced Integrated Intelligence Research, RIKEN s-taiji@is.titech.ac.jp, kanagawa.h.ab@m.titech.ac.jp Hayato Kobayash, Nobuyuki Shimizu, Yukihiro Tagami Yahoo Japan Corporation { hakobaya, nobushim, yutagami } @yahoo-corp.jp
Pseudocode Yes Algorithm 1 Alternating minimization procedure for nonlinear tensor estimation
Open Source Code No The paper does not contain any statement about making its source code publicly available or provide links to a code repository for the described methodology.
Open Datasets Yes Here, we compared the methods in the Restaurant & Consumer Dataset [7]. ... [7] V.-G. Blanca, G.-S. Gabriel, and P.-M. Rafael. Effects of relevant contextual features in the performance of a restaurant recommender system. In Proceedings of 3rd Workshop on Context-Aware Recommender Systems, 2011.
Dataset Splits No Fig. 1 shows the relative MSE (the discrepancy of MSE from the best one) for different training sample sizes n computed on the validation data against the number of iterations t averaged over 10 repetitions. ... The kernel width and the regularization parameter were tuned by cross validation. (The paper mentions 'validation data' and 'cross-validation' but does not provide specific dataset split percentages, sample counts, or detailed splitting methodology needed for full reproduction.)
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., CPU, GPU models, memory, or cloud resources) used to run the experiments.
Software Dependencies No The paper mentions 'Gaussian process method (GP-MTL)' and 'scikit-learn package' (in related works reference), but it does not specify any version numbers for these or other software dependencies.
Experiment Setup No The kernel width and the regularization parameter were tuned by cross validation. (This indicates a tuning strategy but does not provide specific hyperparameter values or detailed training configurations such as learning rates, batch sizes, etc.)