The project concerns the analysis of a financial statement concept catalog as a basis for determining a number of classification problems regarding the filing entity and the content of the financial report from which the information was drawn. The intent of the project is to do a spike inquiry to investigate what methods could be used to correctly classify the companies to an industry segment and if sub-sets of line item concepts catalog could be used to determine specific report components from whcih the set of concepts was draw.
For now this is an educational project aiming to learn the features of the Julia FLuxML deep learning and machine learning library as opposed to the Python PyTorch library. If you are interested in more advanced projects of this nature ... drop me a note at the email below. This is an absolute trivial example just to get familiar with Julia FluxML coding and much more advanced and value projects are already being designed and coded up. Please read my articles on Medium.com to see where I think deep learning, natural language processing, formal logic and computational linguistics will take financial statement analysis and interpretation in the near future.
NOTE: If you want access to the repository with the associated codebase for the project you will soon be able to purchase a NFT crypto-access token that will provide you with the means to peruse the source code in the repository.
For more information contact - Raynier A. van Egmond (XBRL Consulting Partners LLC). Email: raynier@xbrlcp.com LinkedIn: www.linkedin.com\in\rayniervanegmond