功能表聯絡我們登入 Oracle Cloud

Machine Learning in Oracle AI Database

Machine Learning in Oracle AI Database supports data exploration, preparation, and machine learning (ML) modeling at scale using SQL, R, Python, REST, automated machine learning (AutoML), and no-code interfaces. It includes more than 30 high performance in-database algorithms producing models for immediate use in applications. By keeping data in the database, organizations can simplify their overall architecture and maintain data synchronization and security. It enables data scientists and other data professionals to build models quickly by simplifying and automating key elements of the machine learning lifecycle.

Oracle AI Database 26ai 以人工智慧推動資料革命

Oracle 將人工智慧應用於 Oracle AI Database 26ai 的核心架構中，加快實現 Oracle 協助客戶隨時隨地將人工智慧應用於所有資料之承諾。

閱讀新聞稿

Oracle Machine Learning for R updates
Oracle Machine Learning for R now supports Oracle R Distribution 4.4.1, delivering performance gains and enhanced features. It maintains compatibility with open source R 4.4.1 and provides optimized multicore processing and a streamlined developer experience.

Read the news about Oracle Machine Learning for R updates
Prebuilt embedding generation model now available
A downloadable, augmented version of Hugging Face’s all-MiniLM-L12-v2 model in ONNX format is now available. Load it directly into Oracle AI Database and generate vector embeddings with Oracle AI Vector Search.

Learn more about the model about Prebuilt embedding generation model now available
Semantic similarity search with multilingual support
Easily load an augmented version of Hugging Face’s multilingual embedding model, enabling semantic search across 100 languages for crosslinguistic use cases.

Read the blog post about Semantic similarity search with multilingual support

Why choose Machine Learning in Oracle AI Database?

See all features

Oracle AI Database supports data management, model development and deployment options, data and model monitoring, and team collaboration. Enhance productivity through built-in automation, in-database execution performance, and scalability. Identify possible bias in data and understand factors contributing to predictions.

In-database operations

Build models and score data faster and at scale without extracting data to separate analytics engines. Oracle Exadata’s scale-out architecture and Smart Scan technology help deliver results faster.

Multiple language APIs

Choose from SQL, Python, and R interfaces for in-database data exploration and preparation, machine learning modeling, and solution deployment. In addition, deploy Python and R solutions using SQL and REST.

No data movement

Process data where it resides in Oracle AI Database to simplify data exploration and preparation as well as model building and deployment. Shorten application development time, reduce complexity, and address data security.

No-code model building

Improve data scientist productivity and help nonexperts use powerful in-database algorithms for classification and regression through a no-code AutoML user interface.

Data and model monitoring

Gain insights into how your data and machine learning models evolve over time and take corrective action sooner to avoid issues that can have a significant negative impact on the enterprise. Use REST endpoints and no-code user interfaces.

Rapid enterprise deployments

Achieve immediate machine learning model availability with easy deployment options using SQL and REST interfaces.

Bring your own model

Import text transformer, classification, regression, and clustering models in Open Neural Network Exchange (ONNX) format to use from SQL with the in-database ONNX Runtime. Deploy ONNX format models to Oracle Machine Learning Services for real-time inferencing use cases.

High performance compute

Avoid performance issues during data preparation, model building, and data scoring using the built-in parallelism and scalability of Oracle AI Database, with unique optimizations for Oracle Exadata.

Built-in security

Benefit from Oracle AI Database’s built-in security and encryption, role-based access to user data, in-database and third-party models, and R and Python objects and scripts.