Functional Linear Regression of Cumulative Distribution Functions
Authors: Qian Zhang, Anuran Makur, Kamyar Azizzadenesheli
TMLR 2024 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our comprehensive numerical experiments validate the efficacy of our estimation methods in both synthetic and practical settings. Numerical results are displayed in Section 7. |
| Researcher Affiliation | Collaboration | Qian Zhang EMAIL Department of Statistics Purdue University Anuran Makur EMAIL Department of CS and School of ECE Purdue University Kamyar Azizzadenesheli EMAIL Nvidia Corporation |
| Pseudocode | No | The paper describes mathematical models and theoretical results but does not contain any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain any explicit statements about providing source code, nor does it provide links to code repositories or mention code in supplementary materials. |
| Open Datasets | Yes | We compare the empirical performance of our estimator (2) and other methods on two real-world datasets: the California house price dataset (Mohapatra, 2022) and the adult income dataset (Becker & Kohavi, 1996). |
| Dataset Splits | Yes | We split the whole dataset into subsets of fractions 1/3, 1/2, and 1/6. 1/3 data points are used to estimate the coefficients and intercepts under Gaussian or Laplace linear models separately by maximizing log likelihood... Then, we apply different methods on the second subset (training dataset) of 1/2 data points... We calculate L2-errors on the third subset (test dataset) of 1/6 data points for the four methods described previously. |
| Hardware Specification | No | The paper describes numerical experiments in Section 7 but does not provide any specific details about the hardware (e.g., CPU, GPU, memory) used to run these experiments. |
| Software Dependencies | No | The paper mentions "R package stats", "R package L1pack (Osorio & Wolodzko, 2023)", and "R package CVXR (Fu et al., 2020)". However, it does not specify exact version numbers for these packages or the R environment itself, which is required for reproducibility. |
| Experiment Setup | Yes | For the proposed estimator, we calculate pθλ using (3) with S R, m γ0,100, and λ 0.1, 1, 5. For MLE, we can formulate the likelihood function of the parameter θ in (1) with Φ specified in (27). We run the experiments with w 0, 0.5, and 1 in (27). |