Intel® DAAL 2019 is now available. The logistic regression classification algorithm, cross-entropy and logistic loss objective functions are added in the version. Java Maven distribution are added for Java developer. It also boosted machine learning and data analytics performance in EM-GMM, sparse SVM training (linear kernel), logit boost training, gradient boosting prediction and others.
Please Check Intel® DAAL 2019 release notes to learn more information.
What’s New in Intel® DAAL 2019
· Implemented logistic regression classification algorithm.
· Implemented cross-entropy and logistic loss objective functions.
· Added new distribution model via Java MAVEN, which helps JAVA developers to simplify the process of getting and building projects with Intel® DAAL usage.
· Extended traversal API to return additional data from decision tree[KN1] , decision forest, and gradient boosting models.
· Boosted machine learning and data analytics performance in EM-GMM, sparse SVM training (linear kernel), logit boost training, gradient boosting prediction and others.
Deprecation of the following features:
· 32bit library for Mac OS* is deprecated in this release and will be removed in the next release.
Removal of support for:
· Installation on 32-bit hosts for all types of OS is no longer supported. However, the 32-bit library (for Windows* and Linux*) continues to exist, and can be used on 64-bit hosts.
For information about the Intel® DAAL system requirements, please visit the Intel® DAAL 2019 System Requirements page.
Intel® DAAL documentation including Developer Guide and API Reference are available online at Intel® Software Documentation Library. You can also download it from the Intel® Software Development Products Registration Center > Product List > Intel® Parallel Studio XE (or Intel® System Studio) Documentation.
Note: If you encounter a HTTP ERROR: 404 on the beta registration page, please clear the browser cache and try again.