Logo

ML Configurations

Configure Liquid Neural Network parameters and event definitions for fraud detection

Liquid Neural Network Parameters

Configure the core parameters that control how the LNN learns and adapts to fraud patterns

0.001
0.00010.1

Controls how quickly the model adapts to new data

1000
10010000

Number of historical events to remember

0.75
0.11

Minimum confidence score to trigger fraud alert

100
101000

Number of events to analyze in each batch

0.05
0.010.2

How quickly the model adjusts to changing patterns

Event Definitions

Define the structure of events that will be processed by the fraud detection system

Configured Events

payment_transaction

Payment transaction events from payment gateways

Parameters (6)
Requiredamount:number-Transaction amount in cents
Requiredcurrency:string-Transaction currency code
Requiredmerchant_id:string-Unique merchant identifier
Requireduser_id:string-User making the transaction
Optionallocation:string-Geographic location of transaction
Optionaldevice_fingerprint:string-Device fingerprint for fraud detection

login_attempt

User authentication attempts

Parameters (5)
Requireduser_id:string-User attempting to login
Requiredip_address:string-IP address of the attempt
Optionaluser_agent:string-Browser/device information
Requiredsuccess:boolean-Whether login was successful
Requiredtimestamp:string-ISO timestamp of attempt

account_activity

General account activity and changes

Parameters (5)
Requireduser_id:string-User whose account was modified
Requiredaction_type:string-Type of account change
Optionalold_value:string-Previous value before change
Optionalnew_value:string-New value after change
Optionalchanged_by:string-Who made the change

System Status

Current status of the ML configuration system

Configuration Active

ML system is running

5 Parameters

Configured and optimized

3 Event Types

Defined and monitored