├── docsimages ├── importData.png ├── schemaData.png ├── configureRun.png ├── createMLRun.png ├── ml-resource.PNG ├── preparingRUn.png ├── settingsData.png ├── azure-portal-ml.PNG ├── createDataset.png ├── runningModels.png ├── selectDataset.png ├── create-ml-service.PNG ├── resource-created.PNG ├── experimentSettings.png └── finalCreateDataset.png ├── LICENSE └── README.md /docsimages/importData.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/importData.png -------------------------------------------------------------------------------- /docsimages/schemaData.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/schemaData.png -------------------------------------------------------------------------------- /docsimages/configureRun.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/configureRun.png -------------------------------------------------------------------------------- /docsimages/createMLRun.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/createMLRun.png -------------------------------------------------------------------------------- /docsimages/ml-resource.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/ml-resource.PNG -------------------------------------------------------------------------------- /docsimages/preparingRUn.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/preparingRUn.png -------------------------------------------------------------------------------- /docsimages/settingsData.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/settingsData.png -------------------------------------------------------------------------------- /docsimages/azure-portal-ml.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/azure-portal-ml.PNG -------------------------------------------------------------------------------- /docsimages/createDataset.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/createDataset.png -------------------------------------------------------------------------------- /docsimages/runningModels.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/runningModels.png -------------------------------------------------------------------------------- /docsimages/selectDataset.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/selectDataset.png -------------------------------------------------------------------------------- /docsimages/create-ml-service.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/create-ml-service.PNG -------------------------------------------------------------------------------- /docsimages/resource-created.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/resource-created.PNG -------------------------------------------------------------------------------- /docsimages/experimentSettings.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/experimentSettings.png -------------------------------------------------------------------------------- /docsimages/finalCreateDataset.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/GlobalAICommunity/Workshop-VisualInterfaceAutoMLWine/HEAD/docsimages/finalCreateDataset.png -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | MIT License 2 | 3 | Copyright (c) 2019 Global AI Community 4 | 5 | Permission is hereby granted, free of charge, to any person obtaining a copy 6 | of this software and associated documentation files (the "Software"), to deal 7 | in the Software without restriction, including without limitation the rights 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 9 | copies of the Software, and to permit persons to whom the Software is 10 | furnished to do so, subject to the following conditions: 11 | 12 | The above copyright notice and this permission notice shall be included in all 13 | copies or substantial portions of the Software. 14 | 15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 21 | SOFTWARE. 22 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Workshop Automated ML - Predict Wine Quality 2 | 3 | In this case, you want to predict the quality of wine. We used a dataset from UCI Machine Learning repository that contained 6497 observations with physicochemical properties of red and white Portuguese wine and their quality (Cortez). You will use [Azure Machine Learning Automated ML service](https://docs.microsoft.com/en-us/azure/machine-learning/service/concept-automated-ml/?WT.mc_id=gaic-github-cxa). 4 | 5 | ## Steps 6 | 7 | ### Get the data 8 | 9 | You can obtain the data 'winedata.csv' from the GitHub repo [Wine Quality](https://github.com/mdragt/WineQuality) 10 | 11 | ### Create an Azure Machine Learning instance 12 | 13 | Go to [azure.microsoft.com](https://azure.microsoft.com/?WT.mc_id=gaic-github-cxa) 14 | 15 | Sign in and open the Azure Portal (top right) 16 | 17 | Select 'Create a resource' 18 | 19 | Type 'Machine Learning' and 'Create' button 20 | 21 | ![Machine Learning in Azure Portal](docsimages/azure-portal-ml.PNG) 22 | 23 | Enter details: 24 | * **Workspace Name:** provide a name (ex: globalaibootcamp) 25 | * **Subscription:** choose your azure subscription 26 | * **Resource Group:** select new and provide a name (ex: globalaibootcamp) 27 | * **Location:** choose a data centre closest to you 28 | * **Workspace edition:** choose Enterprise in order to use the Automated ML UI 29 | 30 | Select **'Review + Create'** 31 | ![Create Azure Machine Learning Service](docsimages/create-ml-service.PNG) 32 | 33 | Once complete select 'Go to resource' 34 | 35 | ![Go to Resource once created](docsimages/resource-created.PNG) 36 | 37 | Then select 'Launch the new Azure Machine Learning studio' 38 | 39 | ![Launch the new Azure Machine Learning studio](docsimages/ml-resource.PNG) 40 | 41 | This will take you to a new browser window: [https://ml.azure.com/?WT.mc_id=gaic-github-cxa](https://ml.azure.com). Now you have created the Azure Machine Learning instance in the Azure Portal once you can navigate directly to [https://ml.azure.com/?WT.mc_id=gaic-github-cxa](https://ml.azure.com) when you are using the studio. 42 | 43 | 44 | Go to Datasets and select Create dataset, from local file. 45 | 46 | ![Create dataset](docsimages/createDataset.png) 47 | 48 | #### Basic Info 49 | 50 | Select the winedata.csv file. 51 | 52 | ![Select dataset](docsimages/importData.png) 53 | 54 | Click on Next. 55 | 56 | #### Settings and Preview 57 | 58 | Now you have the option to change the settings, but in this case this is not neccesary. 59 | 60 | ![Settings of the dataset](docsimages/settingsData.png) 61 | 62 | Click on Next. 63 | 64 | #### Schema 65 | 66 | The next step is to look at the schema of the data. Here we can change the data types, and deselect columns we don't want in our dataset. In this case, we deselect "quality", which is a numeric value of the wine quality. In this case, we want to predict whether the wine is good or bad, for which we will use the dependent variable "qual_bool". 67 | 68 | ![Schema dataset](docsimages/schemaData.png) 69 | 70 | Click on Next. 71 | 72 | #### Confirm Details 73 | 74 | Finally, you have the option to profile your data, which gives you a generic overview of your data. Therefore, you would need a compute with at least 1 node. 75 | To create the dataset, click on Create. 76 | 77 | ![Confirm dataset](docsimages/finalCreateDataset.png) 78 | 79 | ### Create Experiment 80 | 81 | You are now ready to create a new experiment. Go to Automated ML in the left menu, and select New automated ML run. 82 | 83 | ![Create ML Run](docsimages/createMLRun.png) 84 | 85 | #### Select dataset 86 | 87 | Select the prior created dataset and click on Next. 88 | 89 | ![Select dataset](docsimages/selectDataset.png) 90 | 91 | Fill out the details to create the experiment: 92 | 93 | * Experiment name: any name you want 94 | * Target column: the dependent variable of your dataset, in our case "qual_bool" 95 | * Select training compute target: the compute you want to use for this experiment 96 | 97 | Click on Next. 98 | 99 | ![Experiment settings](docsimages/experimentSettings.png) 100 | 101 | #### Configure Run 102 | 103 | In this step you can configure the ML run. By default, Classification was selected as task, which is ok. Now click on View additional configuration settings. Here you can select things like: 104 | 105 | * Primary metric: define on which metric you want to train the model 106 | * Training job time (in hours): the amount of hours to train the model 107 | * Metric score threshold: minimim metric value that should be reached. I.e. you could be happy with 90% accuracy. 108 | * etc. 109 | 110 | Click on OK, and furthermore on Create. 111 | 112 | ![Configure run](docsimages/configureRun.png) 113 | 114 | The ML run will now be prepared. 115 | 116 | ![Prepare run](docsimages/preparingRUn.png) 117 | 118 | During the run, you get an overview of how well the models are performing: 119 | 120 | ![Running models](docsimages/runningModels.png) 121 | 122 | The running process can take quite a while (remember, we defined 1 hour as possible running time), so take a coffee and go to the next workshop. 123 | 124 | When the running process is finished, you have the option to deploy the best model. 125 | 126 | ## Clean up resources 127 | Finally, If you don't expect to need these resources in the future and want to save cost, you can delete them by deleting the resource group you created. To do so go to the Azure Portal, select 'Resource Groups' and select the resource group for this workshop, select Delete, then confirm the name of the resource group to delete. 128 | 129 | Looking for another example to continue your learning - check out another [tutorial here](https://docs.microsoft.com/en-us/azure/machine-learning/service/tutorial-auto-train-models/?WT.mc_id=gaic-github-cxa) 130 | --------------------------------------------------------------------------------