Are you curious how data scientists and researchers train agents that make decisions? 

Learn how to use reinforcement learning to optimize decision making using Azure Machine Learning.  We show you how to get started.

Time Index:

  • [00:36] – What is reinforcement learning?
  • [01:37] – How do reinforcement learning algorithms work?
  • [04:10] – Reinforcement Learning on Azure – Notebook sample
  • [05:17] – Reinforcement Learning Estimator
  • [07:21] – Sample training Python script
  • [09:06] – Training Result
  • [10:15] – What kind of problems can you solve with reinforcement learning?

Learn More:

The AI Show’s Favorite links:

Building forecasts is an integral part of any business, whether it’s revenue, inventory, sales, or customer demand.

Building machine learning models can be a time-consuming and complex with many factors to consider, such as iterating through algorithms, tuning your hyperparameters and feature engineering.

These choices multiply with time series data, with additional considerations of trends, seasonality, holidays and effectively splitting training data.

Forecasting within automated machine learning (ML) takes these factors into consideration and includes capabilities that improve the accuracy and performance of our recommended models.

This session will highlight the forecasting features of Automated ML and how to leverage them.

Time index:

  • [00:35] – What is time-series forecasting?
  • [01:30] – Simplify ML with Automated ML
  • [02:30] – DriveTime customer scenario
  • [04:15] – Features & Functionality
  • [05:20] – Demo

Learn More:

The AI Show’s Favorite links:

This is Part 2 of a four-part series that breaks up a talk that Seth Juarez gave at the Toronto AI Meetup. (Watch Part 1)

Index:

  • [00:13] Optimization (I explain calculus!!!)
  • [04:40] Gradient descent
  • [06:26] Perceptron (or linear models – we learned what these are in part 1 but I expound a bit more)
  • [07:04] Neural Networks (as an extension to linear models)
  • [09:28] Brief Review of TensorFlow

The Bot Framework Composer is an integrated development tool for developers and multi-disciplinary teams to build bots and conversational experiences with the Microsoft Bot Framework.

In this episode of AI show, Seth Juarez is joined by Vishwac Sena Kannan, Program Manager for Bot Framework to introduce and demo Bot Framework Composer. Visit https://aka.ms/BotFrameworkComp to get started.

Index:
[00:47] – Introduction and overview
[01:45] – Demo – Creating a new bot with Bot Framework Composer
[02:25] – Walkthrough – local bot runtime
[03:30] – Demo – triggers, actions
[05:06] – Language generation integration
[06:08] – Sample bot with Language understanding (LUIS)
[09:00] – Handling interruptions
[11:10] – Wrap up

With ever changing data and customer signals continuous training and retraining can incur higher costs, especially on GPUs for deep learning models.

In AzureML, Microsoft has heard this concern loud and clear, and we want to share some tips to manage your costs and spend your budget wisely.

Come learn about the latest updates to AMLcompute including some tips on saving costs.

Learn More:

Azure Machine Learning has a large library of algorithms from the classification, recommender systems, clustering, anomaly detection, regression and text analytics families.

Each is designed to address a different type of machine learning problem.

In this demo, you will learn how to use Azure Machine Learning designer in a few simple steps and create an end-to-end machine learning pipeline for your data science scenario.

Additional information:

This epsisode of the AI Show talks about the new ML assisted data labeling capability in Azure Machine Learning Studio.

You can create a data labeling project and either label the data yourself, or take help of other domain experts to create labels for you. Multiple labelers can use browser based labeling tools and work in parallel.

As human labelers create labels, an ML model is trained in the background and its output is used to accelerate the data labeling workflow in various ways such as active learning, task clustering, and pre-labeling. Finally, you can export the labels in different formats.

Learn More: