Predict data from a collection to determine a single object with the most used data

I need to start from a set of objects, define a single object, with the most used data among the javascript object collection informed.

Example of set of objects:

[{
    name: "Pedro",
    product: "Caneta"
}, {
    name: "Ana",
    product: "Caderno"
}, {
    name: "Maria",
    product: "Boracha"
}, {
    name: "Pedro",
    product: "Caneta"
}, {
    name: "Pedro",
    product: "Caneta"
}]

Single object with the most used values in the above collections:

{
    name: "Pedro",
    product: "Caneta"
}

That’s what I need, based on the javascript collections array you can see that the most used information for the fields (name and product), were name: "Pedro" and product: "Caneta".
That kind of analysis I need.

What function or how could I do this using tensorflow javascript, if you have any examples I’m very grateful, because I’m new to this.