Index
A
- accuracy, Choosing Evaluation Metrics
- aggregations, Generate features
- algorithm selection, Algorithm Selection
- Apache Spark, Azure Databricks and Apache Spark
- area under curve (AUC), Choosing Evaluation Metrics
- artificial intelligence, The Machine Learning Process
- auto-featurization
- automated machine learning
- on Azure (see Azure Machine Learning service)
- benefits of, Bringing It All Together
- best practices, Best Practices for Machine Learning Projects-Don’t Operate in a Silo
- code-free machine learning, Automated ML for Everyone-Power BI Automated ML to Azure Machine Learning
- development environments, Automated ML for Developers-Conclusion
- feature engineering in (see feature engineering)
- for classification and regression, Using Automated ML for Classification and Regression-Selecting and testing the best model from the experiment run
- growing demand for, Growing Demand
- how it works, How Automated ML Works-End-to-End Model Life-Cycle Management
- iterative process of, An Iterative and Time-Consuming Process-The End-to-End Process
- model deployment, Deploying Automated Machine Learning Models-Conclusion
- model interpretability and transparency, Focus on Transparency to Gain Trust, Enabling Transparency, Feature Engineering and Automated Machine Learning, Model Interpretability and Transparency with Automated ML-Guardrails
- overview of, Machine Learning: Overview and Best Practices-Hyperparameters, Bringing It All Together
- what it is, What Is Automated Machine Learning?-Bringing It All Together
- Azure Active Directory (Azure AD), Azure Databricks and Apache Spark
- Azure Container Instances (ACI), Deploying Models, Web Service Deployment Fails, Web Service Deployment Fails
- Azure Container Services (ACI), Conclusion
- Azure Databricks
- Azure Kubernetes Service (AKS), Deploying to AKS, Web Service Deployment Fails
- Azure Machine Learning service
- Azure Machine Learning software development kit (SDK), Setting Up an Azure Machine Learning Workspace for Automated ML, Azure Notebooks
- Azure Machine Learning workspace
- Azure Notebooks
- Azure portal, Deploying the Model for Testing, Web Service Deployment Fails
- Azure UI, Azure Portal UI-Azure Portal UI
- Azure virtual machines (VMs), Azure Databricks and Apache Spark
C
- categorical features/variables, Feature Engineering and Automated Machine Learning
- class imbalance, What Is Classification and Regression?, Guardrails
- classification
- cloud-based Azure workstations, Notebook VM
- cluster distance, Auto-Featurization for Automated ML
- code examples, obtaining and using, Using Code Examples
- code-free machine learning
- collaboration, Collaboration and Monitoring, Enabling Collaboration
- container images, creating, Creating the Container Image
- ContainerInstance RP, Setting Up an Azure Machine Learning Workspace for Automated ML
- ContainerRegistry RP, Setting Up an Azure Machine Learning Workspace for Automated ML
- Contributor access, Setting Up an Azure Machine Learning Workspace for Automated ML
- cross-validation, Understanding Data
- custom artificial intelligence, The Machine Learning Process
- customer trust, Focus on Transparency to Gain Trust, Model Interpretability and Transparency with Automated ML
D
- data drift, Monitoring and Retraining
- data leakage, Guardrails
- data preprocessing methods, Data Preprocessing Methods Available in Automated ML
- data understanding, Understanding Data
- DataFrames creation, Azure Notebooks
- debugging, Debugging a Deployment
- decision process, understanding, Understand the Decision Process, Classification and Regression
- dependent variables, Machine Learning: A Quick Refresher
- deployment of the model, Deployment
- development environments
- direct explainers, Explainers
- Docker containers, Creating the Container Image
- domain expertise, Feature Engineering and Automated Machine Learning
- driver nodes, Azure Databricks and Apache Spark
E
- embedded feature selection, Select the most impactful features
- encodings, Generate features, Auto-Featurization for Automated ML
- end-to-end (E2E) model life cycle management, End-to-End Model Life-Cycle Management, Setting Up an Azure Machine Learning Workspace for Automated ML
- engineered feature importance, Regression model trained using sklearn
- engineered features, Regression model trained using sklearn
- evaluation metrics, Choosing Evaluation Metrics
- (see also performance metrics)
- experimentation, Embrace Experimentation
- explainers, Explainers
- exploration versus exploitation, Smarter approaches
F
- feature engineering
- auto-featurization, Auto-Featurization for Automated ML-Auto-Featurization for Time-Series Forecasting
- data preprocessing methods, Data Preprocessing Methods Available in Automated ML
- domain expertise and, Feature Engineering and Automated Machine Learning
- example of, Feature Engineering and Automated Machine Learning
- focus points, Feature Engineering and Automated Machine Learning
- importance of, Feature Engineering and Automated Machine Learning
- mastering the art of, Conclusion
- purpose of, Feature Engineering and Automated Machine Learning
- steps of, Feature Engineering, Feature Engineering-Select the most impactful features
- feature generation, Generate features, Auto-Featurization for Automated ML
- feature importance, Model Interpretability, Regression model trained using sklearn
- feature selection, Feature Engineering, Select the most impactful features
- features, Machine Learning: A Quick Refresher
- featuretools package (Python), Conclusion
- filters, Select the most impactful features
- forecasting, Auto-Featurization for Time-Series Forecasting-Auto-Featurization for Time-Series Forecasting
M
- machine learning (see also automated machine learning)
- Machine Learning RPs, Setting Up an Azure Machine Learning Workspace for Automated ML
- mapping business scenarios to data science questions, Understand the Decision Process, Classification and Regression
- MaxAbsScaler, Data Preprocessing Methods Available in Automated ML
- meta explainers, Explainers
- meta-learning, Smarter approaches
- Microsoft Azure Machine Learning service (see Azure Machine Learning service)
- mimic explainer, Explainers
- MinMaxScaler, Data Preprocessing Methods Available in Automated ML
- minority class problem, What Is Classification and Regression?
- missing values, Auto-Featurization for Automated ML, Guardrails
- ML operationalization (MLOps), End-to-End Model Life-Cycle Management
- ML.NET, ML.NET
- model deployment
- model development
- model explainability, Enabling Transparency, Feature Engineering and Automated Machine Learning
- model interpretability, Focus on Transparency to Gain Trust, Enabling Transparency, Model Interpretability-Classification model trained using automated ML
- model life cycle management, End-to-End Model Life-Cycle Management
- model parameters, Model Parameters
- model selection, Selecting a Model-Smarter approaches, Azure Notebooks
- model training process, Hyperparameters
- model transparency, Focus on Transparency to Gain Trust, Enabling Transparency, Feature Engineering and Automated Machine Learning, Model Transparency-Guardrails
- monitoring, Monitoring and Retraining, Collaboration and Monitoring
- multiarmed bandit, Smarter approaches
- multiclass classification, What Is Classification and Regression?
R
- random search, Brute-force approaches
- raw feature importance, Regression model trained using sklearn
- raw features, Regression model trained using sklearn
- regression
- regression pipelines, Registering the Model
- reinforcement learning, Smarter approaches
- remote compute, Azure Databricks and Apache Spark
- resource providers (RPs), Setting Up an Azure Machine Learning Workspace for Automated ML
- retraining, Monitoring and Retraining
- RobustScaler, Data Preprocessing Methods Available in Automated ML
- root-mean-square error (RMSE), Establish Performance Metrics
S
- sampling, Guardrails
- scaling, Data Preprocessing Methods Available in Automated ML
- scoring files, creating, Creating the Container Image
- scoring URI, Deploying the Model for Testing
- search space, Smarter approaches
- search strategy, Smarter approaches
- SHAP Deep Explainer, Explainers
- SHAP Kernel Explainer, Explainers
- SHAP Tree Explainer, Explainers
- skewed data, The Machine Learning Process
- sklearn, Regression model trained using sklearn
- SparseNormalizer, Data Preprocessing Methods Available in Automated ML
- SQL Server, SQL Server
- supervised machine learning, Detecting Tasks
- Swagger documentation, Swagger Documentation for the Web Service
- Synthetic Minority Oversampling Technique (SMOTE), Detect issues with input data and automatically flag them, What Is Classification and Regression?
T
- Tabular Explainer, Explainers
- target encodings, Auto-Featurization for Automated ML
- targets, Machine Learning: A Quick Refresher
- task detection, Detecting Tasks
- Text Explainer, Explainers
- text target encoding, Auto-Featurization for Automated ML
- time-series forecasting, Auto-Featurization for Time-Series Forecasting-Auto-Featurization for Time-Series Forecasting
- transformations, Generate features, Auto-Featurization for Automated ML
- transparency, Focus on Transparency to Gain Trust, Enabling Transparency, Feature Engineering and Automated Machine Learning, Model Transparency-Guardrails
- TruncatedSVDWrapper, Data Preprocessing Methods Available in Automated ML
- trust, Focus on Transparency to Gain Trust, Model Interpretability and Transparency with Automated ML