edit bookmarks

Author: GitHubber17

articles/machine-learning/concept-automl-forecasting-deep-learning.md

@@ -78,7 +78,7 @@ In the table, $n_{\text{input}} = n_{\text{features}} + 1$, the number of predic
 
 ## TCNForecaster in AutoML
 
-TCNForecaster is an optional model in AutoML. To learn how to use it, see [enable deep learning](./how-to-auto-train-forecast.md#enable-deep-learning).
+TCNForecaster is an optional model in AutoML. To learn how to use it, see [enable deep learning](./how-to-auto-train-forecast.md#enable-learning-for-deep-neural-networks).
 
 In this section, we describe how AutoML builds TCNForecaster models with your data, including explanations of data preprocessing, training, and model search. 
 
@@ -104,7 +104,7 @@ The following table lists and describes input settings and parameters for TCNFor
 
 |Training input|Description|Value|
 |--|--|--|
-|Validation data|A portion of data that is held out from training to guide the network optimization and mitigate over fitting.| [Provided by the user](./how-to-auto-train-forecast.md#training-and-validation-data) or automatically created from training data if not provided.|
+|Validation data|A portion of data that is held out from training to guide the network optimization and mitigate over fitting.| [Provided by the user](./how-to-auto-train-forecast.md#prepare-training-and-validation-data) or automatically created from training data if not provided.|
 |Primary metric|Metric computed from median-value forecasts on the validation data at the end of each training epoch; used for early stopping and model selection.|[Chosen by the user](./how-to-auto-train-forecast.md#configure-experiment); normalized root mean squared error or normalized mean absolute error.|
 |Training epochs|Maximum number of epochs to run for network weight optimization.|100; automated early stopping logic may terminate training at a smaller number of epochs. 
 |Early stopping patience|Number of epochs to wait for primary metric improvement before training is stopped.|20|

articles/machine-learning/how-to-auto-train-forecast.md

@@ -127,7 +127,7 @@ Add more detail to this configuration in subsequent sections of this article. In
 
 ---
 
-You specify [validation data](concept-automated-ml.md#training-validation-and-test-data) in a similar way. Create an `MLTable` object and specify a validation data input. Alternatively, if you don't supply validation data, AutoML automatically creates cross-validation splits from your training data to use for model selection. For more information, see the following resources:
+You specify [validation data](concept-automated-ml.md#prepare-training-validation-and-test-data) in a similar way. Create an `MLTable` object and specify a validation data input. Alternatively, if you don't supply validation data, AutoML automatically creates cross-validation splits from your training data to use for model selection. For more information, see the following resources:
 
 - [Select forecasting models](./concept-automl-forecasting-sweeping.md#model-selection-in-automl)
 - [Set training data length requirements](./concept-automl-forecasting-methods.md#data-length-requirements)
@@ -421,7 +421,7 @@ training:
 
 ---
 
-The job searches over all model classes *except* Prophet. For a list of forecasting model names that are accepted in `allowed_training_algorithms` and `blocked_training_algorithms`, see [training properties](reference-automated-ml-forecasting.md#training). You can apply either but not both `allowed_training_algorithms` and `blocked_training_algorithms` to a training run.
+The job searches over all model classes *except* Prophet. For a list of forecasting model names that are accepted in `allowed_training_algorithms` and `blocked_training_algorithms`, see [training properties](reference-automated-ml-forecasting.md#training). You can apply either but not both `allowed_training_algorithms` and `blocked_training_algorithms` to a training run. [](#enable-learning-for-deep-neural-networks)
 
 #### Enable learning for deep neural networks
 

articles/machine-learning/how-to-automl-forecasting-faq.md

@@ -59,7 +59,7 @@ AutoML forecasting supports four basic configurations:
 |Configuration|Scenario|Pros|Cons|
 |--|--|--|--|
 |**Default AutoML**|Recommended if the dataset has a small number of time series that have roughly similar historical behavior.|- Simple to configure from code/SDK or Azure Machine Learning studio. <br><br> - AutoML can learn across different time series because the regression models pool all series together in training. For more information, see [Model grouping](./concept-automl-forecasting-methods.md#model-grouping).|- Regression models might be less accurate if the time series in the training data have divergent behavior. <br> <br> - Time series models might take a long time to train if the training data has a large number of series. For more information, see the [Why is AutoML slow on my data?](#why-is-automl-slow-on-my-data) answer.|
-|**AutoML with deep learning**|Recommended for datasets with more than 1,000 observations and, potentially, numerous time series that exhibit complex patterns. When it's enabled, AutoML will sweep over [temporal convolutional neural network (TCN) models](./concept-automl-forecasting-deep-learning.md#introduction-to-tcnforecaster) during training. For more information, see [Enable deep learning](./how-to-auto-train-forecast.md#enable-deep-learning).|- Simple to configure from code/SDK or Azure Machine Learning studio. <br> <br> - Cross-learning opportunities, because the TCN pools data over all series. <br> <br> - Potentially higher accuracy because of the large capacity of deep neural network (DNN) models. For more information, see [Forecasting models in AutoML](./concept-automl-forecasting-methods.md#forecasting-models-in-automl).|- Training can take much longer because of the complexity of DNN models. <br> <br> - Series with small amounts of history are unlikely to benefit from these models.|
+|**AutoML with deep learning**|Recommended for datasets with more than 1,000 observations and, potentially, numerous time series that exhibit complex patterns. When it's enabled, AutoML will sweep over [temporal convolutional neural network (TCN) models](./concept-automl-forecasting-deep-learning.md#introduction-to-tcnforecaster) during training. For more information, see [Enable deep learning](./how-to-auto-train-forecast.md#enable-learning-for-deep-neural-networks).|- Simple to configure from code/SDK or Azure Machine Learning studio. <br> <br> - Cross-learning opportunities, because the TCN pools data over all series. <br> <br> - Potentially higher accuracy because of the large capacity of deep neural network (DNN) models. For more information, see [Forecasting models in AutoML](./concept-automl-forecasting-methods.md#forecasting-models-in-automl).|- Training can take much longer because of the complexity of DNN models. <br> <br> - Series with small amounts of history are unlikely to benefit from these models.|
 |**Many Models**|Recommended if you need to train and manage a large number of forecasting models in a scalable way. For more information, see the [many models](concept-automl-forecasting-at-scale.md#many-models) article section.|- Scalable. <br> <br> - Potentially higher accuracy when time series have divergent behavior from one another.|- No learning across time series. <br> <br> - You can't configure or run Many Models jobs from Azure Machine Learning studio. Only the code/SDK experience is currently available.|
 |**Hierarchical time series (HTS)**|Recommended if the series in your data have a nested, hierarchical structure, and you need to train or make forecasts at aggregated levels of the hierarchy. For more information, see the [hierarchical time series forecasting](concept-automl-forecasting-at-scale.md#hierarchical-time-series-forecasting) article section.|- Training at aggregated levels can reduce noise in the leaf-node time series and potentially lead to higher-accuracy models. <br> <br> - You can retrieve forecasts for any level of the hierarchy by aggregating or disaggregating forecasts from the training level.|- You need to provide the aggregation level for training. AutoML doesn't currently have an algorithm to find an optimal level.|
 
@@ -77,7 +77,7 @@ AutoML uses machine learning best practices, such as cross-validated model selec
 - **The training data uses features that are not known into the future, up to the forecast horizon**. AutoML's regression models currently assume that all features are known to the forecast horizon. We advise you to explore your data prior to training and remove any feature columns that are known only historically.
 - **There are significant structural differences (regime changes) between the training, validation, or test portions of the data**. For example, consider the effect of the COVID-19 pandemic on demand for almost any good during 2020 and 2021. This is a classic example of a regime change. Overfitting due to regime change is the most challenging problem to address because it's highly scenario dependent and can require deep knowledge to identify.
 
-  As a first line of defense, try to reserve 10 to 20 percent of the total history for validation data or cross-validation data. It isn't always possible to reserve this amount of validation data if the training history is short, but it's a best practice. For more information, see [Training and validation data](./how-to-auto-train-forecast.md#training-and-validation-data).
+  As a first line of defense, try to reserve 10 to 20 percent of the total history for validation data or cross-validation data. It isn't always possible to reserve this amount of validation data if the training history is short, but it's a best practice. For more information, see [Training and validation data](./how-to-auto-train-forecast.md#prepare-training-and-validation-data).
 
 ## What does it mean if my training job achieves perfect validation scores?
 

articles/machine-learning/how-to-use-automated-ml-for-ml-models.md

@@ -313,7 +313,7 @@ If you specified a test dataset or opted for a train/test split during your expe
 > This feature isn't available for the following Automated ML scenarios:
 > - [Computer vision tasks](how-to-auto-train-image-models.md)
 > - [Many models and hiearchical time-series forecasting training (preview)](how-to-auto-train-forecast.md)
-> - [Forecasting tasks where deep learning neural networks (DNN) are enabled](how-to-auto-train-forecast.md#enable-deep-learning)
+> - [Forecasting tasks where deep learning neural networks (DNN) are enabled](how-to-auto-train-forecast.md#enable-learning-for-deep-neural-networks)
 > - [Automated ML jobs from local computes or Azure Databricks clusters](how-to-configure-auto-train.md#compute-to-run-experiment)
 
 Follow these steps to view the test job metrics of the recommended model:
@@ -346,7 +346,7 @@ After your experiment completes, you can test the models Automated ML generates
 > This feature isn't available for the following Automated ML scenarios:
 > - [Computer vision tasks](how-to-auto-train-image-models.md)
 > - [Many models and hiearchical time-series forecasting training (preview)](how-to-auto-train-forecast.md)
-> - [Forecasting tasks where deep learning neural networks (DNN) are enabled](how-to-auto-train-forecast.md#enable-deep-learning)
+> - [Forecasting tasks where deep learning neural networks (DNN) are enabled](how-to-auto-train-forecast.md#enable-learning-for-deep-neural-networks)
 > - [Automated ML jobs from local computes or Azure Databricks clusters](how-to-configure-auto-train.md#compute-to-run-experiment)
 
 If you want to test a different Automated ML generated model, and not the recommended model, follow these steps:

articles/machine-learning/v1/how-to-configure-auto-train.md

@@ -446,7 +446,7 @@ RunDetails(run).show()
 > This feature is not available for the following automated ML scenarios
 >  * [Computer vision tasks](../how-to-auto-train-image-models.md)
 >  * [Many models and hiearchical time series forecasting training (preview)](../how-to-auto-train-forecast.md)
->  * [Forecasting tasks where deep learning neural networks (DNN) are enabled](../how-to-auto-train-forecast.md#enable-deep-learning)
+>  * [Forecasting tasks where deep learning neural networks (DNN) are enabled](../how-to-auto-train-forecast.md#enable-learning-for-deep-neural-networks)
 >  * [Automated ML runs from local computes or Azure Databricks clusters](../how-to-configure-auto-train.md#compute-to-run-experiment)
 
 Passing the `test_data` or `test_size` parameters into the `AutoMLConfig`, automatically triggers a remote test run that uses the provided test data to evaluate the best model that automated ML recommends upon completion of the experiment. This remote test run is done at the end of the experiment, once the best model is determined. See how to [pass test data into your `AutoMLConfig`](how-to-configure-cross-validation-data-splits.md#provide-test-data-preview).