Home

The fundamentals:

• What is a Model?
• What is a Graph?
• To fully understand H1st philosophy and power, check out the H1st Automotive Cybersecurity Tutorial.

Quick start / HelloWorld examples:

• H1st rule-based Model
• H1st ML models
• H1st Graph

Administrative:

• Roadmap
• For Contributors

Productive collaboration is a key value proposition of H1st AI, which provide tools for:

• collaboration between human models (rule-based or fuzzy logic) and machine-learned models
• collaboration between data scientists in a team, each tackling a sub-problem in a decomposed into different graph node

Collaboration tools:

• Graph decomposition: see Graph-Explained and the H1st Automotive Cybersecurity tutorial, chapter 3 for discussion & example of the power of decomposition in action.
• Ensembling: see https://docs.h1st.ai/api/h1st.core.ensemble.html
• Schema & Integration Testing: https://docs.h1st.ai/api/h1st.schema.html
• Model Repository

Model Repository is available out-of-the-box when you subclass from h1.Model API e.g. model.persist() and model.load(). H1st knows how to persist/load Scikit-Learn and Keras models. A simple ML model example with persist/load is available in examples/HelloWorld/ml_model.py.

All you need to do is to set Please set MODEL_REPO_PATH in the project's config.py or calling h1.init(MODEL_REPO_PATH="s3://..."). Model Repository supports local disk or S3 storage.