Lumiata's Lightning AI Model Builder allows Data Scientists, Data Analysts and analytics professionals to train, test, and validate models to publish to production.
Lumiata offers two options for your machine learning needs:
1) Model building using the UI
2) Model building in Jupyter Notebooks
Model Builder in the UI provides a no-code, guided experience to training & testing new ML models. This option is ideal for data enthusiasts who are new to machine learning or for more advanced analytics professionals who are experimenting with new use cases and prefer to get an initial performance baseline in the UI before moving to the Jupyter notebook environment.
Create a new run inside an experiment to build your first ML model. Model Builder will guide you through all the steps, from selecting your target variables and features to configuring the model training and back test time periods. You can also customize advanced settings like train/validate/test splits, sample size, the learning algorithm and hyperparameters.
The JupyterHub integration provides ultimate flexibility for data practitioners comfortable with Python. Get started quickly with our pre-configured notebook templates for EDA, feature engineering, and model training and install any packages you need. Lumiata's proprietary libraries for enrichment, cross-codes, and feature engineering in your notebook save you time so you can focus on building high-performing models.
Find our pre-configured notebook templates for EDA, feature engineering, and model training in the platform here:
To get started, follow our next steps below with the model building option that works best for you.
Updated about 2 months ago