THE COURSE PAPER IN ANALYTICAL CHEMISTRY: PREDICTION OF RETENTION TIME IN HILIC USING THE MACHINE LEARNING METHODS.

In this Jupiter Notebook I'd like to introduce you to my scientific work in Analytical Chemistry (2024-2025), yet in Russian but soon you'll see English desctiption of this particulal notebook.

Description

I've taken the dataset of biological substances from this research and their Experimental Retention Times for building the new ML models and features analysis which are the key for RT modeling.

• I used ensemble and linear models for RT prediction. The LightGBM was the best ($MAE = 0.92, RMSE = 1.43, R^2= 0.75$)

• I've received the Ridge-regresion equation for the TOP-5 desctiptors by their importance:

• I've described the Descriptors.

See more in Coursepaper_finale notebook and have a nice day!!!