What is the data format using Azure machine lerning studio for time series analysis

Question

Hello Azure experts, I want to use Azure machine learning studio to analyze time-series datasets. I used Automated ML and select time series for analyze my data. My data format is like this way, ID for each person, and age at exam is the time stamp points. For example, I want to use the available Weight, Height and available "target predict" value at the different ages to predict the future target_predict data points as shown in "Need to predict".

this format does NOT work for Azure machine learning Automated ML user interface. May I ask what kind of format of time series data format I can use to fit into Azure machine learning Automated ML?

I saw Azure time series insight analysis will be depreciated soon and it is not used for machine learning. So I do not use it.

Answer 1

Hi @Yu Cai

Thank you for your response and confirming that my answer helped.

To reiterate the resolution here, let me jot down the gist of my comment answer above, so that other community users having the same question, can easily reference this!

As per your query, to use Azure Machine Learning Studio for time-series analysis, you can use the Time Series algorithm in the Automated ML interface. However, the data format you provided does not work with the Automated ML user interface.

To fit your data into Azure Machine Learning Automated ML, you can use the following format:

• Each row represents a single observation.
• The first column should be the timestamp.
• The remaining columns should be the features

For instance:

In this example, the timestamp is in the first column, and the remaining columns are the features. You can use the available Weight, Height, and Target_Predict values at different ages to predict the future Target_Predict data points.

I tried to repro with the above dataset and was able to successfully forecast for the above TS data format.

Please refer the doc - How AutoML uses your data for more details.

I hope this helps!

Thank you again for your time and patience throughout this issue.

---

Please don’t forget to Accept Answer and Yes for "was this answer helpful" wherever the information provided helps you, this can be beneficial to other community members.