1

I am new to Bigquery/ Google Cloud services, and my current use-case requires that I am able to generate a training model for each user and then select items from a data table that are the best match for the user. Assuming that the training data is in the same format as the data table I want to retrieve from, how do I go about making this process work with the Prediction API?

In this example, each user has a travel history including how much money they spent at each location, how many nights they stayed, etc.

name, city, hotel, money spent, month/year, nights spent, hotel quality (1-5), personal rating (1-10)
"User1", "San Diego", "Hotel California," 01/2014, 2400, 3, 3, 8
"User1", "Sacramento", "Seedy Stay" 01/2014, 500, 4, 1, 4
"User1", "San Diego", "The Best Hotel", 06/2014, 3000, 4, 5, 10

Now assuming I have a data table of hotels, I would like to be able to predict which hotel(s) would be a good match for User1 from a large data table of hotels. Is this use case possible using just Prediction API and Bigquery or are other steps involved? Or am I overlooking a more straightforward solution?

Thanks in advance!

Erik_01
  • 11
  • 1

1 Answers1

0

By "Prediction API", I presume you mean "Google Prediction API". Please see Redirect Bigquery Data to Prediction.

Community
  • 1
  • 1
Mingyu Zhong
  • 471
  • 4
  • 3