An official Java port of Numenta's Hierarchical Temporal Memory (HTM) library for machine intelligence and anomaly detection.
htm.java is an official Java port of the Numenta Platform for Intelligent Computing (NuPIC), implementing Hierarchical Temporal Memory (HTM) theory. It provides a library for building machine intelligence systems that learn from streaming data, detect anomalies, and make predictions based on temporal patterns. The project maintains close synchronization with NuPIC's core algorithms while offering a Java-native API.
Java developers and researchers interested in biologically-inspired machine learning, anomaly detection in time-series data, and implementing HTM theory in production systems.
Developers choose htm.java for its official community-driven port of NuPIC, offering a performant Java implementation with a 1-to-1 feature correspondence to the original Python library. Its Network API simplifies complex HTM network creation, and it integrates seamlessly with Java build tools like Gradle and Maven.
Hierarchical Temporal Memory implementation in Java - an official Community-Driven Java port of the Numenta Platform for Intelligent Computing (NuPIC).
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Maintains a documented 1-to-1 correspondence with Numenta's NuPIC, ensuring core algorithms like Spatial Pooler and Temporal Memory are accurately implemented as per the project goals.
The Network API reduces configuration complexity for multi-field inference and custom workflows, explicitly highlighted in the README's quick start guide.
Actively maintained with Gitter chat, forums, and contributor calls-to-arms, ensuring ongoing updates and alignment with NuPIC's evolution.
Available via Maven Central and Gradle with no installation required, making dependency management straightforward for Java ecosystems.
Versioned as alpha (e.g., v0.6.13-alpha) with a changelog noting 'unreleased' sections, indicating potential breaking changes and incomplete features unsuitable for production-critical systems.
The versioning table shows Temporal Memory last synced in 2016 versus NuPIC's 2017, admitting temporary lapses that may delay access to latest optimizations or fixes.
Focused solely on HTM theory, it lacks the extensive pre-built models, tutorials, and third-party integrations found in mainstream ML libraries like TensorFlow or scikit-learn.