Browse Prior Art Database

Method and System for a Data-Driven Contest Demand Prediction for Daily Fantasy Sports

IP.com Disclosure Number: IPCOM000246077D
Publication Date: 2016-May-03
Document File: 2 page(s) / 29K

Publishing Venue

The IP.com Prior Art Database

Related People

Maxim Sviridenko: INVENTOR [+4]

Abstract

A method and system is disclosed for predicting demand for contests of daily fantasy sports by using an automated and data-driven technique. The data can be, but need not be limited to, historical or time-related data on how the contests filled up in the past, when the contests were open, when the contests start and if the contests are guaranteed or not. The method and system utilizes information pertaining to features of the contests developed from data insights, modern machine learning tools and architecture implementation. Features are also derived for handling interactions between the various contests in a day.

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 52% of the total text.

Method and System for a Data-Driven Contest Demand Prediction for Daily Fantasy Sports

Abstract

A method and system is disclosed for predicting demand for contests of daily fantasy sports by using an automated and data-driven technique.  The data can be, but need not be limited to, historical or time-related data on how the contests filled up in the past, when the contests were open, when the contests start and if the contests are guaranteed or not.  The method and system utilizes information pertaining to features of the contests developed from data insights, modern machine learning tools and architecture implementation.  Features are also derived for handling interactions between the various contests in a day.

Description

Disclosed is a method and system for predicting demand for contests of daily fantasy sports by using an automated and data-driven technique.  The data can be, but need not be limited to, historical or time-related data on how the contests filled up in the past, when the contests were open, when the contests start and if the contests are guaranteed or not.  The method and system utilizes information pertaining to features of the contests developed from data insights, modern machine learning tools and architecture implementation.  Features are also derived for handling interactions between the various contests in a day.

The method and system selects a set of features that provides the best prediction based on historical data.  For example, a number of real world games and teams playing the real world games correlate with the demand for daily fantasy contests.  As a result, the method and system adds features pertaining to the real world games and the teams to a prediction model.  Thus, the prediction model is trained based on a set of predictive features.

In an embodiment, the method and system uses a Support Vector Machine with Gaussian Kernels as the predict...