# Dig4Bio Raman Transfer Learning Challenge — Data

来源：https://www.kaggle.com/competitions/dig-4-bio-raman-transfer-learning-challenge/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

You will receive:
- Raman spectra from 8 devices (with known concentrations) https://doi.org/10.1016/j.saa.2025.125861 paper from Lange et al. 
- Spectra measured with the new multiplexed system for one full 96-well plate (with known concentrations). There were two spectra recorded for every well of the plate.
- The spectra from a second 96-well plate with unknown concentration.


All spectra available for the challenge are raw spectra without further processing. The spectra of the transfer plate and the 96 spectra were recorded with a Metrohm i-Raman Plus 785 Spectrometer, that features an excitation wavelength of 785nm. The recorded spectra consist of 2048 data points spanning the wavenumber range from 65 cm⁻¹ to 3350 cm⁻¹, with laser power adjustable up to 455 mW to enable fast data acquisition. See further technical details under https://arxiv.org/abs/2504.11234 

## Files

*   **transfer_plate.csv** - the training set
*   **96_samples.csv** - the test set
*   **sample_submission.csv** - a sample submission file in the correct format