You are an ML engineer at a global car manufacture. You need to build an ML model to predict car sales in different cities around the world. Which features or feature crosses should you use to train city-specific relationships between car type and number of sales?
To train city-specific relationships between car type and number of sales, the best approach would involve using feature crosses between the geographical coordinates and car types. Given that latitude and longitude together uniquely define a city location, binning these values and then crossing them separately with the one-hot encoded car type allows for the model to learn specific patterns for each car type in relation to each city’s location. Binning helps manage the variation and scales better, while crossing allows for interactions between the features. Hence, using two feature crosses as an element-wise product: the first between binned latitude and one-hot encoded car type, and the second between binned longitude and one-hot encoded car type makes logical sense.
C https://developers.google.com/machine-learning/crash-course/feature-crosses/check-your-understanding
C should be the answer
D - https://developers.google.com/machine-learning/crash-course/feature-crosses/video-lecture
Cannot be D, Despite Binning is a good idea because it enables the model to learn nonlinear relationships within a single feature; separate latitude and longitude in different feature crosses is not a good one, this separation will prevent the model from learning city-specific sales. A city is the conjunction of latitude and longitude. In that order of Ideas Crossing binned latitude with binned longitude enables the model to learn city-specific effects of car type. I will go for C, https://developers.google.com/machine-learning/crash-course/feature-crosses/check-your-understanding
Damn that was a good explanation. Thank you for writing it out.
C - Answer when doing feature cross the features need to be binned
https://developers.google.com/machine-learning/crash-course/feature-crosses/video-lecture
I got with C
https://developers.google.com/machine-learning/crash-course/feature-crosses/check-your-understanding Answer C: It needs a feature cross to obtain one feature.
C - everything else is madness
creating this cross feature is madness from explainability standpoint
"element-wise product" sounds like we are not using a feature cross but artificially creating a new column whose values is the "element-wise product" of other column values...; i.e., (1, 2, 3) => 1 * 2 * 3 = 6. I am not a native English speaker; thus, I might misunderstand the sentence.
Went with C
C) one feature