Numenta Platform for Intelligent Computing PyTorch libraries

基于智能计算框架nupic的异常检测restful Api.

Thesis project about Unsupervised anomaly detection on the streaming time-series data of porfolio risk measures and returns.

Biologically inspired language model using Jaccard Surprise as its only training signal. No backprop. No GPU. Online Hebbian learning from corrections. Two-layer cortex with apical feedback. Runs on CPU under 200MB.

3D Visualization tool for HTM systems

bare bone Hierarchial Temporal Memory

Task models for human robot collaboration

Simple HTM implementation that uses OpenCL

Indoor localization and pattern recognition using Numenta's HTM framework with ROS

Forward Kinematics of a 3DOF Robotic Arm and Position of End Effector(spherical)

Raspberry Pi 4 Unsupervised Real-Time Anomaly Detection for Streaming Data

State -> Potential Pool -> Permanence -> Activation -> State'

Django Projects for Practice with the Source Code.

This project explores the potential of the Hierarchical Temporal Memory (HTM) algorithm for this purpose, aiming to address challenges such as noise, concept drift, and unknown anomalies. We propose a specialized variant, "Grid HTM," designed to enhance video anomaly detection, making it more adaptive and reliable.

Website for the TEDxNTUA 2019 event

A simple yet powerful Python script that fetches a list of URLs from a text file, converts each web page to a PDF, and merges them into a single

Hierarchical Temporal Memory (HTM) applied to next token prediction.

Cryptographic, Blockchain implemented Online Voting System

Neural style transfer is the process of taking the style of one image And then applying it to the content of another image.