Creating models considering multiple variables

0

Hello: I'm using Forecast to predict Temperature/Humidity/Wind Speed/etc. So far, I'm able to create the models for each parameter individually, but I think the model would benefit considering the other variables. For example, to predict the temperature, it could consider Humidity, Wind Speed, etc to train the model. But I don't see how to do this, as when I import the data, the Schema builder only allows me to include one column of type float, so if I train for Temperature (float type), no other float numbers are allowed (only strings). Is this doable? How should I define the other columns (Humidity, etc.)

Thanks

asked a year ago297 views
1 Answer
0

Hi User,

You should be able to add additional attributes of type float. Here is a sample screenshot of how it can be done: Enter image description here

This is done via the schema builder on AWS Console. Alternatively, it can be done via the JSON Schema like this:

        {
		"AttributeName": "target_value",
		"AttributeType": "float"
	},
	{
		"AttributeName": "humidity ",
		"AttributeType": "float"
	},
	{
		"AttributeName": "temperature",
		"AttributeType": "float"
	},
	{
		"AttributeName": "Wind Speed",
		"AttributeType": "float"
	}

Hope this works for you!

profile pictureAWS
EXPERT
ljunkai
answered a year ago
  • Hello junkai:

    When I try that, I get this error: We were unable to create your dataset. Invalid target time series schema: all non-reserved columns should be of string type. Which is the reason I posted my question...

    Thanks

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions