Summary of Spanseq: Similarity-based Sequence Data Splitting Method For Improved Development and Assessment Of Deep Learning Projects, by Alfred Ferrer Florensa et al.
SpanSeq: Similarity-based sequence data splitting method for improved development and assessment of deep learning projects
by Alfred Ferrer Florensa, Jose Juan Almagro Armenteros, Henrik Nielsen, Frank Møller Aarestrup, Philip Thomas Lanken Conradsen Clausen
First submitted to arxiv on: 22 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Quantitative Methods (q-bio.QM)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This research paper presents a solution to the issue of data leakage in machine learning model assessments for biological sequences. The authors discuss how current practices, such as random splitting of databases into training and testing sets, can produce misleading results due to similarities between samples. They introduce SpanSeq, a method that can scale to large biological sequence datasets to avoid data leakage. The authors also demonstrate the effects of not controlling similarity between sets by reproducing the development of two state-of-the-art models in bioinformatics. This research has implications for the evaluation and improvement of machine learning models in computational biology. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about fixing a problem in how we test computer models that help us understand biological sequences like genes, proteins, and genomes. The current way we do it can be misleading because the data sets are too similar. The researchers created a new method called SpanSeq to fix this issue. They tested two top models for bioinformatics and showed that the old way of testing gives incorrect results. This is important for improving our computer models. |
Keywords
* Artificial intelligence * Machine learning
