Google Professional Machine Learning Engineer
Get started today
Ultimate access to all questions.
Ultimate access to all questions.
You are an ML engineer at a global car manufacturer tasked with building a machine learning model to predict car sales in various cities worldwide. The goal is to capture the relationship between different car types and the number of sales in each city. What features or feature crosses should be used to effectively train the model and capture these city-specific relationships?
Explanation:
The correct answer is C. To effectively capture city-specific relationships between car type and number of sales, feature crosses involving latitude, longitude, and car type are essential. However, creating an element-wise product using raw latitude and longitude would result in a model that might overlook specific geographical granularity. Thus, using binned latitude and binned longitude in combination with one-hot encoded car type provides a more granular and effective representation. This approach enables the model to learn city-specific effects by capturing interactions between location and car types in a more detailed and manageable way.