1 / 9

UW classification: new background rejection trees

This document presents an overview of background rejection trees developed by T. Burnett, detailing the methodology and implementation in merit file structure. Each tree is characterized by a specific set of weighted trees and nodes described in `dtree.txt` and corresponding variables in `variables.txt`. The evaluation process involves using a vector of floats arranged per the variable list. It also discusses training on EVEN events, optional boosting for improved efficiency, and preliminary insights into energy resolution and validity fractions in the context of signal and background weighting.

victor-nicholson
Télécharger la présentation

UW classification: new background rejection trees

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. UW classification:new background rejection trees T. Burnett

  2. The eight selections (from Bill) T. Burnett

  3. The prefilter cuts T. Burnett

  4. Implementation in merit file structure • Each tree is described by two files: • dtree.txt – ascii file with a list of weighted trees and nodes: • tree: specify the weight to assign to the tree • branch: variable index, cut value • leaf: purity • variables.txt – list of the corresponding tuple variables • Evaluation is by passing a vector of floats, ordered according to the variable list. • Proposal to incorporate the prefilter cut in the tree description T. Burnett

  5. Training details • Weight signal and background to be the same • Train on the EVEN events, with optional boosting • Test with ODD events • Save training and testing efficiency curves T. Burnett

  6. Boosting: what does it do? • Nice interpolation for low background • Not much improvement in actual separation (so far) T. Burnett

  7. Preliminary single-tree background T. Burnett

  8. What about the energy resolution?Validity fractions T. Burnett

  9. The fraction of time each estimate is best T. Burnett

More Related