Supported Features and Examples

OML Services extends OML functionality to support data bias detection and data bias mitigation, data monitoring, model monitoring, model deployment, model lifecycle management for both in-database OML models and third-party Open Neural Networks Exchange (ONNX) format machine learning models through REST APIs.

Here are some examples that you might want to perform using the Oracle Machine Learning Services REST API.

Supported Features	Examples
Support for data bias detection and data bias mitigation: Use the data_bias API in OML Services REST API to detect biases in your data, and mitigate the inherent biases.	Oracle Machine Learning Services REST API now supports data bias detection and data bias mitigation. It provides you the data_bias API to detect several types of bias in your data. The data_bias API produces statistics indicating the degree of bias that exists in your data based on well-understood bias metrics. This bias may affect certain groups in the population. Using this API, you can identify bias early in the machine learning model lifecycle. The data_bias API also gives you information to identify several types of data bias with information to help mitigate such bias. For example, you can mitigate bias in your data using reweighing methods to address bias that may have been introduced during the data preparation phase of the machine learning process. For more information, refer to: Work with Data Bias Detection Data Bias Deteciton - Examples
Export in-database models directly to the OML repository: Use OML Services REST API to directly export in-database machine learning models to the OML repository for deployment and scoring.	Oracle Machine Learning Services REST API now allows you to directly export machine learning models from the database to the OML repository for deployment and scoring. The POST/v1/models endpoint also supports json input, where you can provide the name of the in-database model to be exported directly to the repository. For more information, refer to: Work with Oracle Machine Learning Models
Support for Cognitive Image functionality: Use OML Services REST API to work with image classification and to get feature extraction scoring results for input images.	Oracle Machine Learning Services REST API is enhanced to support the following cognitive image functionalities: Image classification along with their respective probabilities Image classification (along with their respective probabilities) of multiple images using the mini batch functionality Image feature extraction results This functionality uses the prebuilt ONNX image model, and is supported by new REST API endpoints: /omlmod/v1/cognitive-image/classification /omlmod/v1/cognitive-image/feature-extraction For more information, refer to: Work with Cognitive Image Functionality Score a Mini Batch Containing base64 Encoded String
Oracle Machine Learning Models: Use OML Services REST API for model deployment and model lifecycle management for both native in-database OML models and third-party Open Neural Networks Exchange (ONNX) machine learning models.	These examples show how to get the list of saved models, get a list of models filtered by model name, version and namespace, store a model, create a model scoring endpoint, get model endpoint details, get model information, get model metadata, get model scoring REST API and so on. For more information, refer to: Deployment Use Cases: Deploy Oracle Machine Learning Models Deploy ONNX Format Models Specifications for ONNX Format Models Examples: Work with Oracle Machine Learning Models Work with ONNX Format Models Work with ONNX Image Models
Batch Scoring: Use OML Services REST API to run batch scoring jobs against Oracle Machine Learning models and ONNX format models on Oracle Autonomous AI Database by using a consistent set of asynchronous REST APIs.	These examples show how to run batch scoring jobs using regression models, classification models, clustering models, and feature extraction models. Work with Batch Scoring Batch Scoring Examples
Data Monitoring: Use OML Services REST API to monitor your data and evaluate how your data evolves over time. It helps you with insights on trends and multivariate dependencies in the data. It also gives you an early warning about data drift.	This example shows how to monitor your data that evolves over time. The example depicts the steps to create and submit data monitoring job, view the job details, and perform various actions such as update, enable, or delete the data monitoring job. Data Monitoring Example
Model Monitoring: Use OML Services REST API to monitor your machine learning models. You can monitor the quality of model predictions over time, and help you with insights on the underlying causes.	This example shows how to monitor your machine learning models. The example depicts the steps to create and submit model monitoring job, view the job details, and perform various actions such as update, enable, or delete the model monitoring job. Model Monitoring Example
Cognitive Text functionality: Cognitive Text functionality, along with Model Repository and Model Deployment, are components of Oracle Machine Learning Services. Oracle Machine Learning Services supports the cognitive text functions through REST API	These examples show how to get a list of model endpoints, return most relevant text keywords, return text summaries, return text sentiment, return most relevant text topics, return text similarity, and return numeric features for text strings. Work with Cognitive Text Functionality