Personalized Cloud Service Scheduling from Predicted User Activity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud computing systems face challenges in managing resources efficiently, as services are often provided 24/7 regardless of user activity, leading to increased costs and resource wastage during inactive periods.
Innovation Solution
A machine learning model trained on signal history data from multiple users predicts user activity levels, enabling personalized service schedules that can pause or disable services during inactive times, using a neural network architecture like the informer for long sequence time-series forecasting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If services are provided 24/7 to ensure availability, then service reliability is improved, but resource consumption and cost increase during inactive periods
Solution Approach 1:
The patent applies dynamics by transitioning from static 24/7 service provision to dynamic service scheduling that adapts to predicted user activity patterns. The system uses machine learning models to forecast when users are likely to be active and adjusts service availability accordingly, making resource allocation flexible and responsive to actual demand patterns rather than maintaining constant availability.
Solution Approach 2:
The patent implements preliminary action by using machine learning models to predict future user activity patterns before actually providing or suspending services. The system analyzes historical data and contextual features to forecast future behavior, allowing proactive scheduling decisions that align service availability with predicted demand rather than reacting to actual usage in real-time.
2Productivity
If personalized service schedules are implemented based on user activity predictions, then resource allocation efficiency is improved, but system complexity increases
Solution Approach 1:
The patent applies copying by using machine learning models that replicate and generalize from historical user behavior patterns to predict future activity. Instead of requiring complex real-time analysis of every individual user's complete activity history, the system creates simplified predictive models that capture essential patterns, allowing efficient personalization without proportionally increasing system complexity.
Solution Approach 2:
The patent implements segmentation by dividing the service scheduling problem into manageable components: collecting historical signal data, extracting contextual features, training predictive models, generating activity predictions, and updating service schedules. This modular approach breaks down the complex task of personalized scheduling into distinct processing stages that can be implemented and maintained more effectively.
3Loss of energy
If services are suspended during predicted inactive periods, then resource consumption is reduced, but service reliability may deteriorate if predictions are inaccurate
Solution Approach 1:
The patent applies feedback by continuously monitoring actual user activity against predicted activity patterns and using this information to refine and retrain the machine learning models. The system updates service schedules based on actual usage patterns, creating a closed-loop system that learns from past performance and improves prediction accuracy over time, thereby reducing the risk of unreliable suspensions.
Data Source
AI summary
The present disclosure relates to providing personalized service schedule in a computing network for a service provider to provide a service. In particular, the systems described herein utilize signal history of a plurality of users to train a model and to predict a cumulative signal amount for an individual user for a predetermined time frame in the future by drawing inferences from the model. The system described herein further transforms the predicted cumulative signal amounts to activity data and a personalized service schedule for the individual user may be updated based on the predicted activity data. The personalized service schedule may be utilized by disabling the service, or pausing the service or pausing a feature of the service when the predicted activity data indicates that the user is likely to be inactive.