Build High-Performance Data Science and Machine Learning Applications

Intel® oneAPI Data Analytics Library (oneDAL) accelerates big data analysis at all stages of the machine learning pipeline. This open source, cross-platform library provides implementations of classical machine learning algorithms that are optimized for Intel CPUs and GPUs.

Using this library, you can:

• Make better predictions faster.
• Optimize data ingestion and algorithmic compute simultaneously.
• Develop offline, streaming, and distributed use models.
• Split analytics workloads between edge devices and cloud to optimize overall application throughput.

oneDAL optimizes algorithms from popular machine learning Python* libraries such as XGBoost and Intel® Extension for Scikit-learn*, which are part of the end-to-end suite of Intel® AI and machine learning development tools and resources.

• To ensure maximum calculation speed, each function is highly tuned to the instruction set, vector width, core count, and memory architecture of each target CPU or GPU.
• See performance benefits for a wide range of applications—from IoT gateways to back-end servers.
• Work in the language you are most familiar with and get maximum performance in your application with integrated Python, SYCL*, and C++ support.

oneDAL includes algorithms for analysis functions, math functions, and training and library prediction functions for C++, including:

• Apriori for Association Rules Mining
• Correlation and Variance-Covariance Matrices
• Decision Forest for Classification and Regression
• Expectation-Maximization Using a Gaussian Mixture Model (EM-GMM)
• Gradient Boosted Trees (GBT) for Classification and Regression
• Alternating Least Squares (ALS) for Collaborative Filtering
• Multinomial Naïve Bayes Classifier
• Multiclass Classification Using a One-Against-One Strategy
• Limited-Memory BFGS (L-BFGS) Optimization Solver
• Logistic Regression with L1 and L2 regularization support
• Linear Regression
• K-Means Clustering
• K-Nearest Neighbor (KNN)
• Support Vector Machines (SVM) with Linear and Radial Basis Function (RBF) Kernels
• Principal Components Analysis (PCA)
• Density-based Special Clustering of Applications with Noise (DBSCAN)
• Random Forest