AI Accelerators Library

Machine learning models have biases using small data, and some industries such as health care and manufaturing lack labled data. In light of this, a good approach is to select robust features to build models.

This accelerator shows how users can apply Local Interpretable Model-agnostic Explanations (LIME) to models built and deployed with DataRobot. LIME serves as another method in your toolbox to explain model predictions, complementing the built-in DataRobot capabilities of XEMP and SHAP prediction explanations.

Modern machine learning techniques are now capable of assisting manufacturers streamline their product development in numerous ways. In this notebook, we are going to focus on detecting and classifying product defects using state of the art computer vision systems. Utilizing machine learning brings immense value to manufacturers, transforming their production processes and giving them an overall competitive edge. By leveraging these advanced methods, manufacturers can streamline product development, enhance defect detection accuracy, optimize operational efficiency, reduce costs, and ultimately deliver higher-quality products to meet the ever-growing demands of the market.

This note book provides a user with a method of changing the output of User Activity Monitor to allow the user to drop an entire column of output or change the contents of that column in a way to preserve the anonymity of the column but maintain consistency for reporting.

This accelerator shows an app to create synthetic training data

This notebook presents an example workflow for carrying out statistical tests, notify stakeholders of any issues via Slack, and generate automated compliance documentation with the test results.

In this notebook, we implement a very simple model based on the Q-learning algorithm. This notebook is intend to show a basic form of Reinforcement Learning that doesn't require a deep understanding of neural networks or advanced mathematics and how one might deploy such a model in DataRobot.

The use case provided in this notebook creates synthetic training data sets for use in DataRobot models. This notebook outlines how to create a synthetic training data set in a csv file, with name, address, phone number, company, account number, and credit score.

This AI Accelerator demonstrates various ways for generating prediction intervals for any DataRobot model. The methods presented here are rooted in the area of conformal inference  (also known as conformal prediction ). These types of approaches have become increasingly popular for uncertainty quantification because they do not require strict distributional assumptions to be met in order to achieve proper coverage (i.e., they only require that the testing data is exchangeable with the training data).

This notebook presents some examples for taking a datarobot project and exporting its model insights as both machine readable files and plots in various file formats.

t-SNE (t-Distributed Stochastic Neighbor Embedding) is a powerful technique for dimensionality reduction that can effectively visualize high-dimensional data in a lower-dimensional space. Dimensionality reduction can improve machine learning results by reducing computational complexity of the algorithms, preventing overfitting, and focusing on the most relevant features in the dataset.

This accelerator aims to assist DataRobot trial users by providing a guided walkthough of the trial experience. DataRobot suggests that you complete the Flight Delays sample use case in the graphical user interface first, and then return to this accelerator.

Understand all of your models in one place and more easily share your findings

Leverage popular python packages as well as DataRobot's python client  to recreate and augment lift chart visualizations.

Leverage DataRobot's python client to extract predictions and compute custom metrics.

Integrate DataRobot API, Papermill, and MLFlow to automate machine learning making it easier, robust, and easy to share.

Access, understand, and tune blueprints for both preprocessing and model hyperparameters.

Learn how to migrate a deployed model using from one DataRobot cluster to another of the same version.

Using Eureqa algorithm to discover the gravitational constant.

Apply FIRE to your dataset and dramatically reduce the number of features.

Import image files using Spark and prepare them into a data frame suitable for ingest

Leverage the power of machine learning to improve customer retention by  building a churn predictor app using Streamlit and DataRobot.

Call the GCP API and enrich a modeling dataset that predicts customer churn

Customize models on the leaderboard via Composable ML's API, the Blueprint Workshop.

Bring external data from Ready Signal to help augment your time series forecasting accuracy

Build model factories leading to the mandatory requirement to significantly decrease training time.

Leverage open source optimization modules to further tune parameters in DataRobot blueprints.