AI Tools Samples Workflow

This workflow is for AI developers looking to explore AI Tools or trying to optimize the performance of their models.

You need access to the AI Tools software. You can download the software to your local development system.

Based on the type of dataset you are using, determine your data type.

• Tabular Data: Non-image data, such as a worksheet of text or numbers that is handled differently from image data.
• Image Data: Specifically visual data, such as images.

Identify where you are in your development lifecycle.

• Data Processing: Convert raw data into a format that libraries can use (for example, pandas) to improve performance.
• Training: Teach a machine learning algorithm using the fit or train function to perform specific tasks (for example, classification or regression).
• Model Optimization: Improve performance by adding additional features or models.
• Inference: Put a model into production.

Identify the specific objective you are working on from one of the following:

• Extract, Transform, Load (ETL): Move data from one or more sources, and then load it to a single model.
• Manipulate Data: Change data to make it easier to process.
• Classification: Supervised learning that categorizes a set of data into different classes.
• Clustering: Unsupervised learning to discover natural groups in the data.
• Regression: Supervised learning to help predict a continuous output variable.
• Dimensionality Reduction: Reduce the number of input variables or features in a dataset.

In the first three steps, you identified your data type, lifecycle stage, and task. The following samples are similarly organized so you can match a sample to your task.

Data Processing : Extract, Transform, Load (ETL)  

• Intel® Distribution of Modin* Get Started Sample
• End-to-End Machine Learning Workload: Census Sample  

Training: Regression  

• Intel® Optimization for XGBoost* Get Started Sample
• Intel® Distribution for Python* Get Started Sample Using daal4py

Inference: Classification  

• Intel Extension for Scikit-learn: Support Vector Classsifier (SVC) for Adult Dataset Sample
• Intel Optimization for XGBoost Get Started Sample
• Intel Distribution for Python Prediction Sample Using daal4py

Inference: Clustering  

• daal4py Distributed K-means Sample

Inference: Regression  

• daal4py Distributed Linear Regression Sample
• End-to-End Machine Learning Workload: Census Sample