Code for Paper On Hypothesis Transfer Learning in Functional Linear Models.
The 'Simulation' Folder contains a synthetic data demo for the proposed TL-FLR.
- './Simulation/data_gen.R' contains the R code for generating synthetic data.
- './Simulation/TL_FLR_funcs.R' contains the R code for TL-FLR/ATL-FLR implementation.
- './Simulation/TL_FLR_demo.R' contains the R code for a demo to compare target-only baseline, TL-FLR, Detect-TL, ALT-FLR (EW), and ATL-FLR.
The stock price data are collected from Yahoo Finance, and we focus on stocks whose corresponding companies have a market cap over
- './Application/Financial-Application/Raw-Data' contains the original stock price data for each sector.
- './Application/Financial-Application/Processed-Data' contains the processed stock price data based on the preprocess description in Appendix G.1.
The dataset is available at https://archive.ics.uci.edu/dataset/240/human+activity+recognition+using+smartphones.
- './Application/HAR-Application/HAR_processed.RData' contains the processed body signal data based on the preprocess description in Appendix G.2.
@misc{lin2024hypothesis,
title={On Hypothesis Transfer Learning of Functional Linear Models},
author={Haotian Lin and Matthew Reimherr},
year={2024},
eprint={2206.04277},
archivePrefix={arXiv},
primaryClass={stat.ML}
}