This folder contains several series of hands-on labs designed to introduce tools and libraries for building intelligent apps that leverage Artificial Intelligence (AI) and machine learning. Each subfolder contains one series of hands-on labs and has a number in the name (200, 300, or 400) that represents the level of technical detail in the series, with 400 being the most advanced. Each series consists of four hands-on labs that build on one another and follow this pattern:
- Ingest - Gather data
- Process - Prepare or clean the data for use in machine learning
- Predict - Train and score a machine-learning model
- Visualize - Visualize the output from the model
To work a series of labs, start with the first lab in the series, then proceed to the second, the third, and the fourth, in that order. Depending on the technical level, a series generally requires 1 to 2 hours to work from start to finish.
| Lab Series | Scenario | Technology/Language | Cost |
|---|---|---|---|
| 200 - Machine Learning in Python | Create a Data Science Virtual Machine DSVM), import a dataset containing on-time arrival information for a major U.S. airline, and use scikit-learn to build a machine-learning model that predicts whether flights will arrive on time. | Data Science Virtual Machine Scikit-learn Jupyter Python |
$$ |
| 300 - Neural Networks with CNTK | Use the MNIST database to train a neural network built with the Microsoft Cognitive Toolkit and Azure Machine Learning Workbench to recognize handwritten digits. Then deploy the model and write a Node.js app to visualize the output. | Microsoft Cognitive Toolkit Azure Machine Learning Workbench Python and JavaScript |
$$ |
| 400 - Image Classification with MMLSpark | Use Bing Image Search to create a database of famous paintings. Then train a deep neural network (DNN) to recognize the artists of those paintings and write a Node.js app that uses the DNN to examine uploaded images and identify the artists. | MMLSpark Azure Machine Learning Workbench Bing Image Search API Azure SQL Database Python and JavaScript |
$$ |
| 400 - Stream Analytics and Machine Learning | Build an end-to-end system that examines photos taken by simulated cameras in the Arctic and identifies those containing pictures of polar bears. Then use Power BI to build a real-time dashboard showing where polar bears are being spotted. | Azure Stream Analytics Azure Machine Learning Workbench Custom Vision Service Azure SQL Database Microsoft Power BI Python and JavaScript |
$$ |
Data science is the new frontier in software development. Use these labs to expand your knowledge of AI, machine learning, and deep learning with neural networks, and acquire first-hand experience with some of the tools of the trade.