Learning from Aggregate responses: Instance Level versus Bag Level Loss Functions

Authors: Adel Javanmard, Lin Chen, Vahab Mirrokni, Ashwinkumar Badanidiyuru, Gang Fu

ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our analysis enables us to theoretically understand the effect of different factors... Additionally, we propose a mechanism for differentially private learning... We also carry out thorough experiments to corroborate our theory and show the efficacy of the interpolating estimator. 5 NUMERICAL EXPERIMENTS Numerical verification of the theory In our first set of experiments, we corroborate our theory derived in Section 2.2 with simulations.
Researcher Affiliation Collaboration 1Google Research, 2University of Southern California
Pseudocode Yes Algorithm 1 Label differentially private learning from aggregate data
Open Source Code No The paper does not provide any concrete access information for source code, such as a specific repository link, an explicit code release statement, or a mention of code in supplementary materials.
Open Datasets Yes Boston Housing dataset. To investigate the optimal value of the regularization parameter ρ in the interpolating loss, for different bag sizes, we conduct numerical experiments on the Boston Housing dataset... (Harrison Jr & Rubinfeld, 1978).
Dataset Splits No The paper uses the Boston Housing dataset and refers to 'test loss' but does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning into train/validation/test sets.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions using a 'feed-forward neural network' and 'Re LU' but does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment.
Experiment Setup Yes We use a feed-forward neural network to learn the housing prices in this dataset. The network has four hidden layers, each with 64 neurons. The activation function for all hidden layers is Re LU. The output layer has one neuron which outputs the predicted housing price.