Repository navigation
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Repository files navigation
Author: Joseph Kaji This goal of this project was to learn how to design my own Neural Network using primarily NumPy. Other significant libraries that were used were pandas for data manaagement and rdkit for scraping NMR. One significant challenge was utilizing the NMReDB2 database as lack of data led to a bottle neck at around 73% accuracy and 0.48 loss. To run this, please download the nmrshiftdb2.nmredata.sd file from the NMReDB2 website and run Make_Model. Then run main.py and enter your sd file with NMR values. Data source for training: https://nmrshiftdb.nmr.uni-koeln.de/ Inspired by: Neural Networks from Scratch