Complex Event Processing for Secure Real-Time Time-Series Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing big data processing systems struggle to efficiently process time-series data sets in real-time to derive meaningful and dynamic predictive data points, especially in environments requiring high data security and privacy.

Innovation Solution

Implementing a cloud-based system with complex event processing engines using recurrent neural networks and Nonlinear Autoregressive Moving-Average (NARMA) models to identify patterns in time-series data, combined with multi-variate regression analysis to derive dynamic predictive data points, leveraging a private cloud platform for enhanced security.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If complex event processing engines are implemented to process time-series data in real-time, then prediction accuracy and processing efficiency are improved, but system complexity and computational resources required increase

Engineering Contradiction:

Improvedata processing efficiencyVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system segments the complex event processing into modular components: data streaming channels for data ingestion, CEP engines for pattern recognition, and machine learning engines for predictive analysis. Each module handles specific tasks independently, reducing overall system complexity while maintaining high processing efficiency.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediate processing layers including event normalization modules and pattern matching intermediaries that bridge raw data input and final predictive output. These intermediaries simplify the processing pipeline by preprocessing data before it reaches the main CEP engines, reducing their computational burden.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If recurrent neural networks with feedback loops are used to account for historical data dependencies, then prediction accuracy is improved, but processing time and computational load increase

Engineering Contradiction:

Improveprediction accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary processing of historical data by pre-computing feature statistics and storing aggregated historical patterns in optimized data structures. This preliminary action reduces the computational burden during real-time prediction, allowing the recurrent neural network to focus only on the most recent data dependencies.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements a hybrid approach where only the most recent and relevant historical data points are processed through the full recurrent neural network with feedback loops, while older data is handled through simplified statistical methods. This partial application of complex processing maintains accuracy for critical recent patterns while reducing overall processing time.

Inventive Principle:

Principle #16Partial or excessive action

3Measurement precision

If multi-variate regression analysis with dynamic weighting factors is implemented, then value variable derivation accuracy is improved, but computational complexity and memory requirements increase

Engineering Contradiction:

Improvevalue variable accuracyVSAvoidmemory requirements

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The system applies local quality by using different weighting strategies for different input variables based on their individual characteristics and relevance. Rather than uniformly processing all variables with complex dynamic weighting, the system identifies and applies simplified weighting to less critical variables while reserving full multi-variate analysis for the most influential factors.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent dynamically adjusts the number and complexity of parameters used in multi-variate regression based on data characteristics and processing requirements. The system can adaptively reduce the dimensionality of input variables when memory constraints are detected, maintaining acceptable accuracy while reducing computational burden.

Inventive Principle:

Principle #35Parameter changes

4Reliability

If data is processed in a private cloud-based platform, then data security and privacy are improved, but system deployment complexity and cost increase

Engineering Contradiction:

Improvedata securityVSAvoiddeployment complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent designs the CEP and machine learning engine architecture to be platform-agnostic and highly portable. The same modular components can be deployed in private cloud, public cloud, or hybrid environments without significant modification, reducing deployment complexity while maintaining the security benefits of private cloud when required.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS12530575B2Intelligent and adaptive complex event processor for a cloud-based platform

Publication Date: 2026.01.20 BANK OF AMERICA CORP

AI summary

Streaming data channels are configured to inject time series-based data sets into a private cloud platform at which complex event processing engines implementing a recurrent neural network using a Nonlinear Autoregressive Moving-Average (NARMA) model identifies patterns in the data sets to continuously determine future/predictive data points for each data set. The dynamic future/predictive data points/values serve as inputs to machine learning engines, which implement multi variate regression analysis to derive a dynamic variable value for a mechanism.