Patent license prediction using deep survival analysis: A comparative study

Takao Arai, Hiroyasu Inoue · 2026

Read Patent license prediction using deep survival analysis: A comparative study by Takao Arai, Hiroyasu Inoue (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0355826.txt.

Description

Takao Arai and Hiroyasu Inoue published a 2026 study in PLOS ONE on predicting which patents will be licensed. They use survival analysis, the statistics of time until an event, rather than a simple classifier. They compared old and new methods on over six hundred thousand American patents: classical regression, a plain neural network, a neural survival model, and fancier cure-rate versions assuming some patents never license.

All the deep learning approaches performed strongly, far ahead of classical regression at spotting the patents that did license. The standard neural survival model gave the best balance of simplicity and accuracy. The cure-rate extensions added almost nothing in practice, despite one statistically detectable edge, and survival framing itself beat static classification only modestly. The real gains, they conclude, come from deep learning's grip on complex patent features, not from modeling time or hidden subgroups, so users can keep the simpler model.

Catalog record

CREATORTakao Arai, Hiroyasu Inoue
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE46781
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0355826

Browse the full library · About the Tower of Babel