Leading the New Era of Machine Intelligence
Numenta has developed a cohesive theory, core software technology, and numerous applications all based on principles of the neocortex. This technology lays the groundwork for the new era of machine intelligence. Our innovative work delivers breakthrough capabilities and demonstrates that a computing approach based on biological learning principles will make possible a new generation of capabilities not possible with today’s programmed computers.
Biological and Machine Intelligence (BAMI)BAMI is a living book that documents Hierarchical Temporal Memory, a theoretical framework for both biological and machine intelligence.
Read it here: http://numenta.com/biological-and-machine-intelligence/
What Is Machine Intelligence?
Because today’s computers are programmed, they can only do exactly as they are told. In stark contrast, intelligent machines continuously and automatically learn patterns in their environment without being programmed, enabling them to tackle problems in entirely new ways. Intelligent machines that learn and act will have an enormous beneficial impact in the coming decades.
Brain as a Blueprint
The brain is the best example of an intelligent system and provides a roadmap for building intelligent machines. The brain’s center of intelligence, the neocortex, controls a wide range of functions using a common set of principles. Numenta has made significant progress discovering these principles and using them to create learning algorithms. Numenta is unique in its understanding and adherence to neocortical principles.
We’ve created the first benchmark designed to evaluate real-time anomaly detection algorithms. The Numenta Anomaly Benchmark is an open-source framework, with real-world data files, that rewards early detection and offers a controlled method for measuring and comparing performance.
Learn more about NAB.
Our machine intelligence technology is called Hierarchical Temporal Memory (HTM), which is a detailed computational theory of the neocortex. At the core of HTM are time-based learning algorithms that store and recall spatial and temporal patterns. HTM is well suited to a wide variety of problems, particularly those with the following characteristics:
- Streaming data rather than static databases
- Underlying patterns in the data change over time
- Many individual data sources where hand crafting separate models is impractical
- Subtle patterns that can’t always be seen by humans
- Time-based patterns
- Simple techniques such as thresholds yield substantial false positives and false negatives
Our technology has been tested and implemented in software, all of which is developed with best practices and is suitable for deploying in commercial applications. Our core learning algorithms are fully documented and available in an open source project called NuPIC.
Like the brain, Numenta’s machine intelligence technology can be applied to many types of problems. We have tested it on a variety of applications to validate its broad applicability. A few of these applications are described below.
HTM for Stocks
Grok for IT Analytics
Rogue Behavior Detection
Models normal behavior of individuals. Detects changes in behavior indicative of unauthorized file access or unauthorized trading. You can experiment with this application using your own data by downloading our sample application code below.
Detects anomalies in the movement of people, objects, or material using speed and location data. Enables logistics optimization. You can experiment with this application using your own data by downloading our sample application code below.
Numenta Anomaly Benchmark (NAB)
Early anomaly detection in streaming data is as difficult as it is important. Yet no tools exist for comparing detection techniques for real-time, streaming data.
That’s why we’ve created the Numenta Anomaly Benchmark (NAB). NAB is an open source framework that anyone can use to test and compare real-time anomaly detection algorithms. It consists of:
- A dataset with 58 real-world, labeled data files
- A scoring mechanism that rewards early detection and on-line learning
- Business Paper: The Numenta Anomaly Benchmark
- Technical Peer-Reviewed Paper: Evaluating Real-time Anomaly Detection Algorithms – the Numenta Anomaly Benchmark
- More Information
Interested in exploring our technology or developing your own application? Our machine intelligence algorithms, encoders, and application code are all available on NuPIC (Numenta Platform for Intelligent Computing), our open source project.
Numenta aims to make it easy for business to develop and distribute applications that harness the power of our machine intelligence technology. Learn about licensing opportunities below or contact Numenta to inquire about potential partnerships.
Earlier this month, I attended the annual Computational and Systems Neuroscience meeting (Cosyne) in Salt Lake CityRead More
Massachusetts General Hospital recently established a center in Boston that plans to use Nvidia chips to help an arRead More
- Wed, Jun 01, 2016 7:00 PM — 8:00 PM
- OC Deep Learning, HTM, ANN, NLP & AI MeetupIrvine, CA USA
We receive many requests for better documentation of our work, and as our research has progressed, we know that somRead More