Using Azure Machine Learning to Predict Who Will Survive the Titanic - Part 2

Download this episode

Download Video

Description

In this session, you will learn to use Azure Machine Learning to make predictions.  The example used is predicting whether a passenger on the Titanic will survive, given information like their age, gender, class of ticket, ticket fare, etc.  But these same principles can be used to predict if someone will make a purchase online or whether a patient will be readmitted to a hospital in the next 30 days.

Previously in Part 1, Jennifer Marsman has already demonstrated how to upload a dataset into Azure Machine Learning Studio, explore the data and decide how to modify it, and use data cleaning modules to implement these changes.  In this Part 2, we train a model with a machine learning algorithm, deploy our model, and call our published model to get results. 

1:14: Split the data

4:00: Choose a machine learning algorithm

9:02: Implement in Azure Machine Learning Studio

12:50: Evaluate the models

16:14: Deploy a model

18:31: Call a webservice to access our published model 

 

Embed

Format

Available formats for this video:

Actual format may change based on video formats available and browser capability.

    The Discussion

    • User profile image
      Geoffers

      Excellent primer for Azure ML. Any chance of more videos? 

      Thanks very much  Jennifer.

       

    • User profile image
      Jeff

      Very cool video, thank you.
      I wish we could have seen a demonstration on scoring the test set for Kaggle.
      I also have one doubt/question. I believe you called "Scored Probabilities" a confidence level. But it seems that is more of a score that could be manipulated to change Accuracy and Precision.
      That seems like another available tool to change models based on penalties of false positive or false negative.

    Comments closed

    Comments have been closed since this content was published more than 30 days ago, but if you'd like to send us feedback you can Contact Us.