# Working with experiments

Load historical data into an automated machine learning experiment and train a model to analyze and predict a business problem.

You can create and edit experiments in personal or shared spaces.

## Workflow

Before you create an automated machine learning experiment in Qlik Cloud Analytics, you need to have a well-defined machine learning question and a suitable training dataset available in Catalog. For more information, see [Defining machine learning questions](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/machine-learning-framework.htm) and [Getting your dataset ready for training](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/dataset-preparation.htm).

The following steps describe an experiment workflow.

1. Create your experiment

   Create a new experiment in Qlik Cloud. Add it to a shared space if you want to work collaboratively.

   [Creating experiments](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/creating-experiments.htm)

2. Configure your experiment

   Select a target to make predictions on and features to support the prediction.

   [Configuring experiments](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/configuring-experiments.htm)

3. Start the training

   Start the training of your first experiment version.

   [Training experiments](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/training-experiments.htm)

4. Review recommended models

   Review how the model training went and assess the models that have been recommended to you. Use the built-in tools to analyze training summaries and embedded analytics, which offer insights at the experiment, version, and model level.

   [Reviewing models](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/reviewing-experiments.htm)

   If further refinement is needed, you can configure subsequent versions of the experiment. Adjust parameters such as features and algorithms and retrain new versions of the experiment until you have a good model.

   [Refining models](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/refining-models.htm)

5. Deploy the model

   When you have a good model, it’s time to deploy it and start making predictions.

   [Working with ML deployments](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/working-with-deployments.htm)

## Requirements and permissions

For information about the required space roles for working with ML experiments in shared spaces, see:

* [Shared space permissions for Professional and capacity-based subscription users](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/Spaces/permissions-shared-space-prof-full.htm)

* [Shared space permissions for administrators](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/Spaces/permissions-shared-space-admin.htm)

If you are an administrator, see [Who can work with Qlik Predict](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/who-can-work-with-automl.htm) for a comprehensive overview of the required user permissions for working with ML experiments.

## Viewing lineage and impact analysis

Using the Lineage and Impact analysis tools in Qlik Cloud, you can analyze:

* The origins of an ML experiment and the data used to train each model.

  [Analyzing lineage for machine learning content](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/Catalog/lineage.htm#MLLineage)

* How the ML experiment and its outputs are being used in downstream content across Qlik Cloud.

  [Impact analysis for machine learning content](https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/Catalog/impact-analysis2.htm#MLImpact)