Hierarchical Temporal Memory and Semantic Folding in Node.JS
$12-25 USD / hour
Ακυρώθηκε
Αναρτήθηκε πάνω από 6 χρόνια πριν
$12-25 USD / hour
The ideal candidate has strong Node + Math + Machine learning experience, as we have to build this solution entirely in Node.JS ES6 from scratch.
Implementation Order & Methodology:
1. Implement a fast and efficient HTM in Node.JS from scratch
[login to view URL]
2. Identify and prepare a training corpus for step 3.
3. Apply semantic folding to your HTM implementation
[login to view URL]
4. Implement semantic fingerprinting (SF) routines in Node:
+ create sparse binary vectors employing two-dimensional topographic semantic space
+ calculate distance between semantic spaces
5. Implement our cutting-edge applications on top of the Node SF routines.
Alternatively: train NN model in Tensorflow, vector operations in Node.
Further references:
[login to view URL]
[login to view URL]
[login to view URL]
Try demos of such implementation in action:
[login to view URL]
[login to view URL]