LeStrat-Net is a machine learning algorithm designed to enhance Monte Carlo simulations through a novel stratification approach. Traditional Monte Carlo methods divide the domain space of the integrand into regular intervals, but LeStrat-Net uses a different strategy based on the height of the function being sampled, akin to Lebesgue integration. This means that isocontours of the function define regions that can have any shape, depending on the function's behavior. Neural networks are employed to learn these complex functions and predict the divisions, allowing for preclassification of large samples of the domain space. This preclassification enables tasks such as variance reduction, integration, and event selection to be performed more efficiently. The network defines the regions it has learned and is also used to calculate the multi-dimensional volume of each region, offering a more flexible and efficient approach to Monte Carlo simulations.
Neural networks, stratification
Not specified
Not specified
Variance reduction, integration accuracy
Not specified
No
No
Flexible stratification, neural network-based learning
No
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
No
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
Not specified
No
Not specified
Not specified
No
Not specified
Not specified
Not specified
Not specified
Not specified
No
Not specified
Not specified
0.00
Not specified
Not specified
01/01/1970
01/01/1970
Not specified
Not specified
Yes