Dataset and code for: [Machine Learning-based Investigation of the Relationship Between Plasma Emission Spectra and Biological Responses]

This dataset and code supports the study titled: "Machine Learning-based Investigation of the Relationship Between Plasma Emission Spectra and Biological Responses."

This research focuses on developing and validating machine learning models to predict biological outcomes (24-hour and 48-hour MTT cell viability assays) from the optical emission spectra and operational parameters of two custom-built helium plasma jets.

The uploaded file, Plasma Prj.rar, contains all the necessary components to reproduce the results of this study. The contents are organized as follows:

1. Codes

This folder contains all Python scripts used for the entire workflow, including:

• Data preprocessing and normalization.

• Training, validation, and testing of all machine learning models (e.g., Random Forest, CNN, etc.).

• Final analysis and figure generation.

2. DataSets

This folder contains the datasets from both plasma jets, divided into raw and processed versions.

• A. Raw Data: This subfolder contains the original, unprocessed data files for both plasma devices.

• B. Final DataSets (Processed and Normalized): This subfolder contains the final, analysis-ready CSV files:

  • train_processed.csv: The training partition (n=520) from the primary plasma jet (Jet 1).

  • test_processed.csv: The hold-out test partition (n=130) from the primary plasma jet (Jet 1).

  • processed_second_jet.csv: The complete, processed dataset for the external validation device (Jet 2, n=62).