Language with NuPIC
- Python modules enum, pandas, and numpy.
- You must have a valid REST API key from Cortical.io.
In this directory are scripts for some NLP experiments. They each use one or more of the following models:
- Keywords model gives a random (seeded) encoding to each word.
- Cortical.io word fingerprints model encodes each word with their API, and then makes a union representation over the words in a block of text.
- Cortical.io document fingerprints model uses their API to encode a block of text.
- HTM Network models use the Cortical.io word-level encoder, and then run the patterns through the Network API.
- Cortical.io classify endpoint model is an end-to-end classification tool in their API.
Please refer to the Imbu project directory.
The "Fox-eats" demo is an example from the NuPIC 2013 Hackathon, using a previous version Numenta's NLP code.