AutoML Tables for model where comparison is required?

[spiderbetuk]

Hi there,

I have used GCP for a while now, and have trained quite a few models using AutoML Tables - all of these have been fairly simple datasets with probably a maximum of 20 columns.

I now have a problem that I would like to solve, but the dataset is a lot more complicated. I want to be able to predict the results of Greyhound Racing, or at least the % chances of each Greyhound winning a given race, compared to the other greyhounds running in that same race.

To be able to do this I need to feed multiple pieces of data for each Greyhound in each given race, to be able to predict the winning chance of that greyhound in that day's race.

However, I am very stuck on how to structure my data. Using AutoML Tables - would I need to structure the data in a tabular form with many columns? Or is there a better way to tackle this problem.

Here is an example of the data I would be using:

Race:

Example data for each Greyhound in the race:

Does anyone please have any advice of how to tackle this kind of problem, and how best to structure the data to attempt to predict the winning chance of each Greyhound for that day's race, based on that greyhound's previous data, compared to the other greyhounds in that day's race?

Thanks,
Rob

[SeaportAi]

Hi, this is tabular data and hence has to be structured as a tabular data. I know this will call for lot of data prep work.