scicloj.ml
An idiomatic Clojure machine learning library providing a unified interface for classification, regression, and unsupervised models.
What is scicloj.ml?
scicloj.ml is an idiomatic Clojure machine learning library that provides a unified interface for various classification, regression, and unsupervised models. It solves the problem of fragmented ML tooling in Clojure by offering harmonized, data-driven pipelines and sophisticated cross-validation.
Clojure developers and data scientists looking for an idiomatic, functional approach to machine learning with seamless integration into Clojure workflows.
Developers choose scicloj.ml for its idiomatic Clojure design, extensible plugin architecture, and efficient data handling via established libraries like tech.ml.dataset and Smile.
Overview
A Clojure machine learning library
Use Cases
Best For
- Building machine learning pipelines as data in Clojure
- Performing cross-validation on ML models in a Clojure environment
- Integrating Smile or XGBoost models into Clojure applications
- Extending ML capabilities with plugins for non-JVM languages
- Preprocessing datasets efficiently using tech.ml.dataset
- Experiment tracking via callback mechanisms in ML workflows
Not Ideal For
- Teams working exclusively outside the JVM or Clojure ecosystem, as it requires Clojure-specific knowledge and setup.
- New projects requiring a stable, long-term API with active maintenance, given the deprecation of this shim in favor of 'noj' or direct library usage.
- Developers seeking extensive built-in deep learning models without relying on external plugins or integrations.
- Projects needing a simplified, all-in-one ML solution without managing multiple underlying libraries like tech.ml.dataset and Smile.
Pros & Cons
Pros
Provides a unified, functional interface that integrates seamlessly with Clojure workflows, enabling data-driven pipelines as first-class citizens, as emphasized in the philosophy.
Leverages tech.ml.dataset for high-performance data storage and manipulation, ensuring scalability and speed with large datasets, as noted in the key features.
Open architecture allows easy integration of models from non-JVM languages, including deep learning, via plugins like sklearn.clj and scicloj.ml.clj-djl.
Includes built-in, easy-to-use cross-validation tools for evaluating pipeline performance, which is highlighted as a main feature for model selection.
Cons
The library is marked as deprecated and recommends using underlying libraries directly or the new 'noj' library, which may confuse users and limit future updates or support.
Acts as a shim over multiple libraries (e.g., tech.ml.dataset, Smile, metamorph), adding complexity that can obscure debugging and increase learning overhead.
Documentation is spread across external tutorials and repositories, and the deprecation note makes it challenging to find reliable, up-to-date guidance for new users.
Frequently Asked Questions
Related Projects
PyTorch - Tensors and Dynamic neural networks in Python with strong GPU accelerationTensors and Dynamic neural networks in Python with strong GPU acceleration
kerasDeep Learning for humans
streamlitStreamlit — A faster way to build and share data apps.
gradioBuild and share delightful machine learning apps, all in Python. 🌟 Star to support our work!
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.