Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]
Contract Scheduling with Predictions
Authors: Spyros Angelopoulos, Shahin Kamali
JAIR 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | For both prediction models, we complement the theoretical analysis with an experimental evaluation of our schedules. The results demonstrate that the empirical improvements are in line with the theoretical analysis. |
| Researcher Affiliation | Academia | Spyros Angelopoulos EMAIL CNRS and LIP6 Sorbonne University 4 place Jussieu, Paris, France 75252 Shahin Kamali EMAIL Department of Electrical Engineering and Computer Science York University, Toronto, Canada |
| Pseudocode | No | The paper describes algorithms and schedules using mathematical notation and textual descriptions (e.g., "Consider the geometric schedule G = (bi r) i=1"), but it does not contain any clearly labeled pseudocode blocks or algorithm boxes with structured, code-like formatting. |
| Open Source Code | Yes | The code on which the experiments are based is available online at https://github.com/shahink84/Contract-Schduling-With-Predictions. |
| Open Datasets | No | The paper describes a theoretical framework for 'contract scheduling' and evaluates its proposed 'schedules.' The 'data' for experiments is generated (e.g., 'we generate 1,000 predictions associated with t') rather than using external, publicly available datasets. |
| Dataset Splits | No | The paper does not use any external datasets, so there is no mention of training/test/validation splits. The experimental evaluation involves generating predictions and random errors: 'For each discrete value of t (as discussed in more detail later in the section), we generate 1,000 predictions associated with t, in accordance with the two models we study.' |
| Hardware Specification | No | The paper does not provide any specific details regarding the hardware (e.g., GPU models, CPU types, memory) used to run the experiments. It only states that code is available online. |
| Software Dependencies | No | The paper does not specify any software dependencies (e.g., programming languages, libraries, frameworks, solvers) along with their version numbers that would be needed to reproduce the experiments. |
| Experiment Setup | Yes | We model τ [T H, T + H] to be a random, normal variable with mean T and standard deviation 1, such that η H. ... We run the experiment over 1,000 evenly spaced values of the interruption time in the interval [2, 220]. For each value of the interruption time, the expectation is taken over 1,000 random values of the error. |