We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Computer Science > Machine Learning

Title: Direct-Manipulation Visualization of Deep Networks

Abstract: The recent successes of deep learning have led to a wave of interest from non-experts. Gaining an understanding of this technology, however, is difficult. While the theory is important, it is also helpful for novices to develop an intuitive feel for the effect of different hyperparameters and structural variations. We describe TensorFlow Playground, an interactive, open sourced visualization that allows users to experiment via direct manipulation rather than coding, enabling them to quickly build an intuition about neural nets.
Subjects: Machine Learning (cs.LG); Human-Computer Interaction (cs.HC); Machine Learning (stat.ML)
Cite as: arXiv:1708.03788 [cs.LG]
  (or arXiv:1708.03788v1 [cs.LG] for this version)

Submission history

From: Daniel Smilkov [view email]
[v1] Sat, 12 Aug 2017 15:36:26 GMT (6288kb,D)

Link back to: arXiv, form interface, contact.