Model Choice for Machine Learning RRS feed

  • Question

  • I have a list of Prices and associated information from which a future Price, let's say 1 hour in the future, should be able to be predicted.

    What is the best choice for a Learning Model in this case?

    How do I choose a Learning Model for this case?

    How do I build a Learning Model for this case?

    Tuesday, September 25, 2018 2:51 PM

All replies

  • It's a time series forecasting project, so you must focus on specific algorithms. There is a great book authored by Microsoft seniors from which you can get valuable information about this: Practical Time Series Analysis, Dr. Avishek Pal and Dr. PKS Prakash.

    Here also you can find a summary of the most important algorithms:

    7 methods to perform Time Series forecasting (with Python codes)

    Good luck

    Tuesday, September 25, 2018 4:33 PM
  • Where do I find a Time Series Analysis model in Microsoft's Azure Machine Learning Studio?

    Tuesday, September 25, 2018 6:42 PM
  • There is just one built-in module in AML Studio for Time Series, but customized for Anomaly Detection.

    Time Series Anomaly Detection

    You should consider to use Python (or R) scripts within your experiment, and write your own algorithms using libraries like statsmodels and of course pandas and sklearn.

    If this was helpful, don't forget to mark the reply as answer, to help others with similar concerns.

    Tuesday, September 25, 2018 9:01 PM
  • Is there a Microsoft Azure Machine Learning Studio experiment that shows sample usage of  Time Series Anomaly Detection module?

    How do you visualize the output of the Time Series Anomaly Detection module?
    Tuesday, September 25, 2018 10:03 PM