Oracle adds machine learning features to MySQL HeatWave

Oracle is adding new machine learning features to its MySQL HeatWave data analytics cloud service.

MySQL HeatWave combines OLAP (Online Analytical Processing), OLTP (Online Transaction Processing), machine learning, and AI-driven automation in a single MySQL database.

The new machine learning capabilities will be added to the service’s AutoML and MySQL Autopilot components, the company said when it announced the update on Thursday.

While AutoML enables developers and data analysts to build, train, and deploy machine learning models within MySQL HeatWave without moving to a separate service for machine learning, MySQL Autopilot provides machine learning-based automation to HeatWave and OLTP, such as automatic provisioning, automatic encoding, query plan, automatic shape prediction, and automatic data location, among other features.

AutoML increases time series forecasting through machine learning

New machine learning-based capabilities added to AutoML include multivariate time series forecasting, unsupervised anomaly detection and recommender systems, Oracle said, adding that all the new features were generally available.

“Multivariate time series forecasting can predict multiple variables ordered in time, where each variable depends on both its past value and the past values ​​of other dependent variables. For example, it is used to build forecast models to predict winter electricity demand by taking into account the various energy sources used to generate electricity,” said Nipun Agarwal, Oracle’s senior vice president of research.

Unlike the usual practice of having a statistician trained in time series analysis or forecasting select the correct algorithm for the desired outcome, AutoML Multivariate Time Series Forecasting automatically processes the data to select the best algorithm for the ML model. and automatically adjusts the model, the company said.

“HeatWave AutoML’s automated forecasting pipeline uses a proprietary technique consisting of stages including advanced time-series preprocessing, algorithm selection, and hyperparameter tuning,” Agarwal said, adding that this automation can help companies save time and effort, as they do not need to have trained statisticians on staff.

According to Holger Muller, Principal Analyst at Constellation Research, the multivariate time series forecasting feature is unique to Oracle’s MySQL HeatWave.

“Time series forecasting, multivariate or otherwise, is not currently offered as part of a single database that offers analytics augmented by machine learning. AWS, for example, offers a separate database for time series,” Muller said.

HeatWave improves anomaly detection

Along with multivariate time series forecasting, Oracle is adding machine learning-based “unsupervised” anomaly detection to MySQL HeatWave.

In contrast to the practice of using specific algorithms to detect specific anomalies in data, AutoML can detect different types of anomalies from unlabeled data sets, the company said, adding that this feature helps business users when they don’t know what types of anomalies are in the data set

“The model generated by HeatWave AutoML provides high accuracy for all types of anomalies: local, cluster and global. The process is fully automated, eliminating the need for data analysts to manually determine which algorithm to use, which features to select, and optimal hyperparameter values,” Agarwal said.

In addition, AutoML added a recommendation engine, which it calls recommender systems, which supports automation for algorithm selection, feature selection, and hyperparameter optimization within MySQL HeatWave.

“With MySQL HeatWave, users can call the ML_TRAIN procedure, which automatically trains the model which is then stored in the MODEL_CATALOG. To predict a recommendation, users can call ML_PREDICT_ROW or ML_PREDICT_TABLE,” Agarwal said.

Business users get MySQL HeatWave AutoML console

Additionally, Oracle is adding an interactive console for business users within HeatWave.

“The new interactive console allows business analysts to create, train, run and explain ML models using the visual interface, without using SQL commands or any coding,” Agarwal said, adding that the console makes it easy for business users to explore scenarios. conditional for your company

“The addition of the interactive console is in line with companies trying to make machine learning responsible. The console will help business users dive into the deeper end of the pool as they want to become ‘citizen data scientists’ to avoid getting into too much hot water,” said Tony Baer, ​​Principal Analyst at dbInsight.

The console is initially available for MySQL HeatWave on AWS.

Oracle also said it would add Amazon S3 storage support for HeatWave on AWS to reduce costs and improve service availability.

“When data is loaded from MySQL (InnoDB storage engine) into HeatWave, a copy is made to the scale-out data management layer built into S3. When an operation requires data to be reloaded into HeatWave, such as during failover, the data can be accessed by multiple HeatWave nodes in parallel and the data can be loaded directly into HeatWave without the need for any transformation,” Agarwal said.

MySQL Autopilot Updates

New features added to MySQL HeatWave include two new additions to MySQL Autopilot: automatic prediction advisor integration with the interactive console and automatic download.

“Within the interactive console, database users can now access the MySQL Autopilot Auto shape prediction advisor that continuously monitors the OLTP workload to recommend with explanation the correct compute shape at any given time. , which allows customers to always obtain the best price-performance ratio”. Agarwal said.

The automatic download feature, according to the company, can recommend which tables to download based on workload history.

“Freeing up memory reduces the size of the cluster required to run a workload and saves cost,” Agarwal said, adding that both features were generally available.

HeatWave targets smaller data volumes

Oracle offers a smaller form of HeatWave to attract customers with smaller data sizes.

In contrast to the previous 512GB size for a standard HeatWave node, the smallest form will be 32GB in size with the ability to process up to 50GB for a price of $16 per month, the company said.

Additionally, the company said that the data throughput capacity for its standard 512GB HeatWave node has increased from 800GB to 1TB.

“With this increase and other improvements in query performance, HeatWave’s price performance benefit has further increased by 15%,” said Agarwal.