Online Ad Allocation with Predictions

Authors: Fabian Spaeh, Alina Ene

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We experimentally evaluate our algorithm on synthetic and real-world data on a wide range of predictions. Our algorithm is consistently outperforming the worst-case algorithm without predictions.
Researcher Affiliation Academia Fabian Spaeh Department of Computer Science Boston University fspaeh@bu.edu Alina Ene Department of Computer Science Boston University aene@bu.edu
Pseudocode Yes Algorithm 1 Exponential Averaging with Predictions
Open Source Code No The paper does not provide any statement or link regarding the public availability of its source code.
Open Datasets Yes We generate two instances for Display Ads based on the real-word datasets i Pin You (Zhang et al., 2014) and Yahoo (Yahoo, 2011).
Dataset Splits No The paper does not specify explicit training, validation, or test dataset splits (e.g., percentages or sample counts) for reproducibility, as the problem is online with impressions arriving sequentially.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., CPU, GPU models, memory) used for running the experiments.
Software Dependencies No The paper does not specify software dependencies with version numbers (e.g., Python, PyTorch, specific solver versions) needed to replicate the experiments.
Experiment Setup Yes For each predictor, we show the consistency (left) and robustness (right) for varying α. Figure 4 shows results for α = 5 with predictions of different quality, as described in the figure caption. We vary the sample fraction ϵ [0, 1] for the dual base algorithm and p [0, 1] for random and biased corruptions. We use synthetic data with 12 advertisers and 2000 impressions of 10 types, where we report the same quantities as in Figure 3. We assume a constant budget for each advertiser of 10 impressions as it makes for an interesting instance.