Path9.Mod1.a - Selecting Regression Algorithms for Azure ML Flashcards
Augmented learning https://learn.microsoft.com/en-us/azure/machine-learning/algorithm-cheat-sheet?WT.mc_id=docs-article-lazzeri&view=azureml-api-1 https://learn.microsoft.com/en-us/azure/machine-learning/media/algorithm-cheat-sheet/machine-learning-algorithm-cheat-sheet.png?view=azureml-api-1#lightbox
- Designer supports two types of Components; Classic Prebuilt (v1) and Custom Components (v2), which are compatible (T/F)
- New projects should use Custom Components (T/F)
- False. These Component types are not compatible.
- True. They are compatible to AzureML v2 and will continue to receive new updates. v1 Classic Prebuilt won’t receive new functionality.
When you want to Predict values with a fast training linear model
Linear Regression
When you want to Predict Event Counts
Poisson Regression
When you want to Predict values with linear models on small datasets
Bayesian Linear Regression. Bayesian Algorithms inherentily incorporate Normalization (a form of Regularization), making them less likely to overfit on smaller datasets.
When you want to Predict a Distribution or understand the distribution of values, rather than just the value
Fast Forest Quantile Regression
When you want to Predict values with high accuracy and fast training times, while maintaining control over memory utilization
Decision Forest Regression
When you want to Predict values with high accuracy, where lengthy training times are not a concern
Neural Network Regression
When you want to Predict values with high accuracy and fast training times, where memory utilization is not a concern
Boosted Decision Tree Regression
