Classify events based on brem or not 2 classes of events Brem, with - - PowerPoint PPT Presentation

classify events based on brem or not
SMART_READER_LITE
LIVE PREVIEW

Classify events based on brem or not 2 classes of events Brem, with - - PowerPoint PPT Presentation

Classify events based on brem or not 2 classes of events Brem, with recoil energy < 1500 MeV No brem See if we can distinguish them, two cases should have different shower features Compare 3D CNN, Graph Neural Network 100 000 events for


slide-1
SLIDE 1

Classify events based on brem or not

2 classes of events

Brem, with recoil energy < 1500 MeV No brem

See if we can distinguish them, two cases should have different shower features Compare 3D CNN, Graph Neural Network 100 000 events for training+ validation 100 000 events for testing/predictions

1

slide-2
SLIDE 2

3D CNN

Architecture

1 conv layer, 32 filters of size (8,8,8) 1 max pool of size (5,5,5) 1 dense layer Dropout 0.16

Adam optimizer with learning rate 0.004 ReLu Activation function after conv and dense layer

Max(0,x)

Input size = (45,70,35) 96.4 % accuracy

2

slide-3
SLIDE 3

Graph Neural Network

Architecure

3 EdgeConv blocks 1 deep layer 256 nodes Dropout 0.1

Adam optimizer with learning 0.003 100 hits as input vertices ReLu activation function 83.0% accuracy

3

slide-4
SLIDE 4

ROC curves, CNN & Graph

4

slide-5
SLIDE 5

Mean and Std Dev

5

slide-6
SLIDE 6

Mean and Std Dev Rates

6

slide-7
SLIDE 7

Mean and Std Dev Rates

7

slide-8
SLIDE 8

Mean and Std Dev

8

slide-9
SLIDE 9

Mean and Std Dev

9

slide-10
SLIDE 10

Mean and Std Dev

10

slide-11
SLIDE 11

Energy

11

slide-12
SLIDE 12

Azimuthal Angle

12

slide-13
SLIDE 13

Azimuthal Angle

13

slide-14
SLIDE 14

Energy vs Hits

14

slide-15
SLIDE 15

Energy & Hits

15

slide-16
SLIDE 16

Energy vs Angle

16

slide-17
SLIDE 17

Energy vs Angle

17