A Perceptron-Like Linear Supervised Algorithm for Text Classification
Abstract
A fast and accurate linear supervised algorithm is presented which compares favorably to other state of the art algorithms over several real data collections on the problem of text categorization. Although it has been already presented in [6], no proof of its convergence is given. From the geometric intuition of the algorithm it is evident that it is not a Perceptron or a gradient descent algorithm thus an algebraic proof of its convergence is provided in the case of linearly separable classes. Additionally we present experimental results on many standard text classification datasets and artificially generated linearly separable datasets. The proposed algorithm is very simple to use and easy to implement and it can be used in any domain without any modification on the data or parameter estimation.
Type
Publication
In Advanced Data Mining and Applications (ADMA), Lecture Notes in Computer Science, 2010, Volume 6440/2010, 86-97, Chongqing, China
Authors
Senior AI/ML and genomics practitioner
Senior AI/ML and genomics practitioner with ~15 years building open-source, production-grade tools for large-scale biological data. Maintainer of multiple Bioconductor packages (fastreeR, metabinR, jvecfor), and author of agentic, LLM-driven tooling that runs reproducible bioinformatics workflows from natural-language requests.
Broad multi-omics background spanning genome assembly and annotation, population genomics, large-scale NGS and functional-genomics analysis, and metagenomics, backed by reproducible HPC software and end-to-end program leadership. Currently focused on bringing modern AI — embeddings, deep learning, and LLM-based agents — to making complex omics datasets faster and easier to interrogate.