Machine Learning Ensemble for Sales Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods lack accuracy in predicting future sales and transaction metrics in dealership environments, such as automobile sales, due to limitations in data utilization and model training processes.
Innovation Solution
A method involving multiple machine-learning models (MLMs) trained on diverse datasets with different input variables, combining predictions for enhanced accuracy, and a data visualizer for presenting current and predicted metrics to users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single machine-learning model is used for prediction, then the model complexity is low, but the prediction accuracy is insufficient
Solution Approach 1:
The patent divides the prediction system into multiple separate machine-learning models, each trained on different subsets of input variables. This segmentation allows each model to specialize in specific data patterns, improving overall prediction accuracy while maintaining manageable complexity through modular architecture
Solution Approach 2:
The patent combines predictions from multiple machine-learning models by averaging their outputs. This merging approach integrates diverse predictive insights from different models, achieving higher accuracy than any single model could provide while distributing the computational complexity across multiple simpler components
2Loss of information
If all input variables are used in a single model, then data utilization is high, but the model becomes difficult to train and interpret
Solution Approach 1:
The patent segments the full set of input variables into different subsets, assigning each subset to a separate machine-learning model. This segmentation enables efficient training of individual models on manageable data portions while collectively utilizing all available input data across the ensemble of models
Solution Approach 2:
The patent applies local quality by giving different models different input variable subsets based on their specific training purposes. Each model optimizes its performance on particular data characteristics, allowing the system to effectively utilize diverse data types without overwhelming any single model's training capacity
Data Source
AI summary
A first outcome variable to be predicted by a machine-learning model (MLM) is determined. The first outcome variable is associated with a product. Using product information that comprises values for each of a plurality of different input variables, a plurality of MLMs are trained to predict the first outcome variable, each MLM utilizing a different set of input variables of the plurality of different input variables. Using historical data that identifies historical values for the first outcome variable, each MLM is tested to determine an accuracy for each MLM. A first MLM is identified based on the testing.