Today, participants began working on the first problem of the individual round. The dataset for the problem was prepared by the Avito team.
The participants were asked to imagine that they are developers who need to help a marketplace automatically calculate shipping costs. To do this, the model had to predict an item's actual parameters based on the listing text and a photo of the item:
The more accurately the model determines these characteristics, the more correctly the shipping cost is calculated — and the fewer surprises at the pickup point. In simple terms: the task is to teach a neural network to "understand" whether it's looking at a guitar, a suitcase, or a board game, and what dimensions they have.
The ranking for today's round is already available at: cups.online/ru/rounds/1470/leaderboard
Tomorrow, participants face the second problem of the individual round — we promise it will be just as interesting!
The participants were asked to imagine that they are developers who need to help a marketplace automatically calculate shipping costs. To do this, the model had to predict an item's actual parameters based on the listing text and a photo of the item:
- weight,
- height,
- width,
- length.
The more accurately the model determines these characteristics, the more correctly the shipping cost is calculated — and the fewer surprises at the pickup point. In simple terms: the task is to teach a neural network to "understand" whether it's looking at a guitar, a suitcase, or a board game, and what dimensions they have.
The ranking for today's round is already available at: cups.online/ru/rounds/1470/leaderboard
Tomorrow, participants face the second problem of the individual round — we promise it will be just as interesting!