Multidimensional Time-Series Data for Software Activity Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine learning models trained using one-dimensional time-series data struggle to accurately predict software application activity due to noise and failure to recognize seasonal patterns, leading to inefficient allocation of computing resources.
Innovation Solution
Generating multidimensional time-series data organized across various time granularities to train machine learning models, allowing for more accurate prediction of activity patterns and efficient resource allocation by differentiating between temporary and permanent changes, and accounting for seasonality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are trained using one-dimensional time-series data, then the training process is simple, but the prediction accuracy is low due to noise and inability to recognize seasonal patterns
Solution Approach 1:
The patent transforms one-dimensional time-series data into two-dimensional data structures by introducing multiple time granularities (e.g., daily, weekly, monthly aggregations). This dimensional transformation enables the model to capture seasonal patterns and trends while maintaining manageable complexity through structured aggregation of temporal data at different scales.
2Productivity
If computing resources are allocated based on inaccurate predictions, then resource allocation is inefficient, but improving prediction accuracy requires complex multidimensional data processing
Solution Approach 1:
The patent segments time-series data into multiple granularities (e.g., hourly, daily, weekly) and processes each segment separately. This segmentation allows the system to handle complex temporal patterns through manageable chunks, improving both prediction accuracy and resource allocation efficiency without overwhelming computational complexity.
Solution Approach 2:
By adding the time granularity dimension to the data structure, the patent enables more accurate prediction of software application activity. This additional dimension allows the model to distinguish between temporary fluctuations and seasonal patterns, leading to better resource allocation decisions that balance accuracy requirements with processing complexity.
Data Source
AI summary
Certain aspects of the present disclosure provide techniques for predicting activity within a software application using a machine learning model. An example method generally includes generating a multidimensional time-series data set from time-series data associated with activity within a software application. The multidimensional time-series data set generally includes the time-series data organized based on a plurality of time granularities. Using a machine learning model and the generated multidimensional time-series data set, activity within the software application is predicted for one or more time granularities of the plurality of time granularities. Computing resources are allocated to execute operations using the software application based on the predicted activity within the software application.