/onlinehd

Efficient single-pass hyperdimensional classifier. Mirror of https://gitlab.com/biaslab/onlinehd

Primary LanguageJupyter Notebook

hyperdimension

This project with UCI researchers deals with hyperdimension learning

Envs

MacOS. 11.2.3

Ubuntu. 20.04 - GPU Server(RTX3060)

Python : 3.8.5

Torch : 1.9.0

==========================================================================

01_example.ipynb

  • You can confirm input data type and return data type

02_Encoder.ipynb

  • You can confirm OnlineHD Encoder

03_HD_Example2.ipynb

  • Hyperdimension computing encoding and decoding example.

04_Mnist.ipynb

  • You can confirm hypervector and learning result of MNIST dataset.

05_isolet.ipynb

  • OnlineHD Training with ioslet dataset.

06_OnlineHD_search.ipynb

  • OnlineHD Encoding research.

07_hyper-dimension.ipynb

  • Mapping Hypervector in 2-D and 3-D through t-SNE random label.

08_hyper-dimension.ipynb

  • Mapping Hypervector in 2-D and 3-D through t-SNE label 0-5, 5-9.

09_custom_encoding_onlinehd.ipynb

  • Custom OnlineHD and confirm learning result and making decoder.

10_HDGenerator.ipynb

  • Making Hyperdimension Generator