Designing and Implementing a Data Science solution on Azure -koulutuksessa käydään läpi uudet tavat tehdä Machine Learning -malleja Azuressa ilman koodausta Machine Learning Designerilla, sekä Notebookeilla.
Lisäksi koulutuksessa käydään läpi koko ML Pipeline: Datan käsittely, laskentaympäristön pystyttäminen, algoritimien automatisoitu valinta, hyperparametrien virittäminen, tuotantoon siirto, monitorointi jne. Siis käytännössä koko ketju alusta loppuun modernin koneoppimisen toteuttamiseen Azure-ympäristössä.
Tavoite
Opi kuinka teet modernia koneoppimista Azuressa.
Kenelle
Tämä koulutus on suunnattu data- ja IT -asiantuntijarooleille mm. Data Scientist, Data Engineer, Data Analyst, IT-Architect, IT-Specialist ja Applications Specialist.
Koulutus edellyttää Python-kielen perusteiden sekä Machine Learning -perusteiden tuntemusta esimerkiksi Azure AI Fundamentals -kokonaisuutta tai vastaavia taitoja.
Lisätiedot
Koulutus valmentaa Microsoftin viralliseen Exam DP-100 Designing and Implementing a Data Science Solution on Azure -sertifiointitestiin.
Koulutuksen sisältö
Design a machine learning solution
- Determine the appropriate compute specifications for a training workload
- Describe model deployment requirements
- Select which development approach to use to build or train a model
Manage an Azure Machine Learning workspace
- Create an Azure Machine Learning workspace
- Manage a workspace by using developer tools for workspace interaction
- Set up Git integration for source control
- Create and manage registries
Manage data in an Azure Machine Learning workspace
- Select Azure Storage resources
- Register and maintain datastores
- Create and manage data assets
Manage compute for experiments in Azure Machine Learning
- Create compute targets for experiments and training
- Select an environment for a machine learning use case
- Configure attached compute resources, including Azure Synapse Spark pools and serverless Spark compute
- Monitor compute utilization
Explore data by using data assets and data stores
- Access and wrangle data during interactive development
- Wrangle interactive data with attached Synapse Spark pools and serverless Spark compute
Create models by using the Azure Machine Learning designer
- Create a training pipeline
- Consume data assets from the designer
- Use custom code components in designer
- Evaluate the model, including responsible AI guidelines
Use automated machine learning to explore optimal models
- Use automated machine learning for tabular data
- Use automated machine learning for computer vision
- Use automated machine learning for natural language processing
- Select and understand training options, including preprocessing and algorithms
- Evaluate an automated machine learning run, including responsible AI guidelines
Use notebooks for custom model training
- Develop code by using a compute instance
- Track model training by using MLflow
- Evaluate a model
- Train a model by using Python SDK v2
- Use the terminal to configure a compute instance
Tune hyperparameters with Azure Machine Learning
- Select a sampling method
- Define the search space
- Define the primary metric
- Define early termination options
Run model training scripts
- Configure job run settings for a script
- Configure compute for a job run
- Consume data from a data asset in a job
- Run a script as a job by using Azure Machine Learning
- Use MLflow to log metrics from a job run
- Use logs to troubleshoot job run errors
- Configure an environment for a job run
- Define parameters for a job
Implement training pipelines
- Create a pipeline
- Pass data between steps in a pipeline
- Run and schedule a pipeline
- Monitor pipeline runs
- Create custom components
- Use component-based pipelines
Manage models in Azure Machine Learning
- Describe MLflow model output
- Identify an appropriate framework to package a model
- Assess a model by using responsible AI principles
Deploy a model
- Configure settings for online deployment
- Configure compute for a batch deployment
- Deploy a model to an online endpoint
- Deploy a model to a batch endpoint
- Test an online deployed service
- Invoke the batch endpoint to start a batch scoring job
Apply machine learning operations (MLOps) practices
- Trigger an Azure Machine Learning job, including from Azure DevOps or GitHub
- Automate model retraining based on new data additions or data changes
- Define event-based retraining triggers
Avainsanat
Azure Machine Learning, Data Science