This video talks about Azure Machine Learning Pipelines, the end-to-end job orchestrator optimized for machine learning workloads. With Azure ML Pipelines, all the steps involved in the data scientist's lifecycle can be stitched together in a single pipeline improving inner-loop agility, collaboration, and reuse of data and code, while maintaining high reliability. Learn More: Azure ML Pipeline sample notebooksWhat are Azure ML PipelinesCreate and run your first PipelineThe AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning Get Started with Machine Learning

Building accurate machine learning models requires a deep understanding of machine learning, as well as other data science and coding skills. Automated machine learning's purpose is to alleviate these requirements, automating the ML workflow of data cleansing, feature engineering, algorithm selection, hyperparameter tuning, and deployment. Enabling business domain experts to leverage ML in their day to day work using the web UI without writing a single line of code, as well as optimizing data scientists work by automating part of the technical work. Learn More: Automated ML Docs How to work with automated ML UINotebook Samples Blog Post [00:55] - Start an Automated ML Experiment [03:07] - Analyze experiment outputs and deploy recommended model The AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning Get Started with Machine Learning

In this episode we'll cover a quick overview of new capabilities in Azure Machine Learning like the Visual Interface for drag-n-drop machine learning and the Automated Machine Learning UI as well as our MLOps approach to productionalize machine learning from training to deployment. We'll also quickly highlight how to prepare your data using our new integration with Azure Data Factory and even visualize results of model predictions via our integration with Power BI. Learn More: Try it today for free AML OverviewTry out sample Azure Machine Learning Jupyter notebooksLearn more about MLOps [00:40]- Data preparation via integration with Azure Data Factory[01:53] – Finding new data to improve your model training with Azure Open DataSets[03:00] – Drag n drop machine learning via the Visual Interface[04:10] – New Azure Machine Learning Studio UI[05:45] – Automated Machine Learning in hosted Jupyter Notebook[09:40] – Automated Machine Learning in Azure Machine Learning Studio UI[10:20] – Dataset profiling and exploration in Azure Machine Learning Studio UI[11:11] – Tensorflow deep learning model training in hosted Jupyter Notebook[13:54] – Tensorboard integration[14:20] – Automated hyperparameter sweep via Azure Machine Learning Hyperdrive [14:48] – Easily deploying models to test & production environments with Azure DevOps integration[16:45] – Integrating deployed ML models into Power BI The AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning Get Started with Machine Learning

This video talks about Azure Machine Learning designer that will help citizen data scientists who were using the Studio (classic) and professional data scientists who are currently using Azure ML SDK to build end to end machine learning pipelines. Learn More: Designer overviewTutorialModule reference Batch inference in the designer The AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning Get Started with Machine Learning

In this episode we will discuss the new R capabilities coming to public preview in Azure Machine Learning, including the new Azure ML R SDK. Learn More: Azure ML SDK for R Interface to the Azure ML SDK for R [03:10] - Demo The AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning Get Started with Machine Learning

We'll be discussing updates to the Azure Machine learning service model registry to provide more insights about your model. Also, learn how you can deploy your models easily without going through the effort of creating additional driver and configuration files. Learn More: Learn more about Azure Machine Learning service Try out sample machine learning notebooks[00:00] Intro to Azure Machine Learning model registry [01:50] Intro to Simplified deployment with No code deployment feature [03:01] Demo for Workspace [03:50] Demo for No code deployment The AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning Get Started with Machine Learning

In this episode we'll learn how to build a "Visual Alert" for your IoT camera. We will use Custom Vision from Azure Cognitive Services to train an image classifier for a particular camera, and export that ONNX model to run locally on an IoT device. Our device will send events back to Azure IoT Hub. The end to end code is available on GitHub for your to try along at home HERE. Learn More: Visual Intelligence Docs: Cognitive Services Vision Solution Templates Product Page Iot Visual Alerts [01:16] - Visual Alerts Demo Intro The AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning Get Started with Machine Learning

Understanding what your AI models are doing is super important both from a functional as well as ethical aspects. In this episode we will discuss what it means to develop AI in a transparent way. Mehrnoosh introduces an awesome interpretability toolkit which enables you to use different state-of-the-art interpretability methods to explain your models decisions. By using this toolkit during the training phase of the AI development cycle, you can use the interpretability output of a model to verify hypotheses and build trust with stakeholders. You can also use the insights for debugging, validating model behavior, and to check for bias. The toolkit can even be used at inference time to explain the predictions of a deployed model to the end users. Learn more: Link to the docLink to the sample notebooksSegments of the video: [02:12] – Responsible AI[02:34] – Machine Learning Interpretability[03:12] – Interpretability Use Cases[05:20] - Different Interpretability Techniques[06:45] - DemoThe AI Show's Favorite links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning

In this episode Luis stopped by and showed how much more can really be done with Cognitive Search (with recipes to boot). Extracting structure from unstructured data is a powerful addition to Cognitive Search! The demo he presents gives an amazing step-by-step process for using Cognitive search to enrich your index. Main Demo: [04:56] Learn More About Knowledge Mining: Main Page for Azure Search: http://aka.ms/cognitivesearch Solution Accelerator: http://aka.ms/kmsolutions Azure Search DocumentationThe AI Show's Favorite Links: Don't miss new episodes, subscribe to the AI Show Create a Free account (Azure) Follow Seth on Twitter AI Blog Fast ML MIT News | AI Medium | Francesca Lazzeri Deep Learning vs. Machine Learning

Machine Learning is sometimes confusing. From the esoteric terms to elevated expositions it seems like a terribly difficult area to get into. Since I started as a developer I totally get the mismatch! In this episode we tackle the one term that is used all of the time in Machine Learning: the elusive "model." First we set up how machine learning is different, how to think about it, and finally what a model actually is (spoiler alert - think "a function written a different way"). Would love your feedback! Learn More: Machine Learning Models The AI Show's Favorite Links: Don't miss new episodes, subscribe to the AI ShowCreate a Free account (Azure)Follow Seth on TwitterAI BlogFast MLMIT News | AIMedium | Francesca LazzeriDeep Learning vs. Machine Learning