Deep Learning Variable Importance for Pharmaceutical Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pharmaceutical claims processing systems rely on static data models, which are inaccurate and prone to errors due to evolving trends in pharmaceutical data, making it difficult to predict trends and price changes in prescription drug markets effectively.
Innovation Solution
A machine learning system utilizing a combination of deep learning methods, including long-short term memory (LSTM) and multilayer perceptron (MLP) algorithms, along with predictive artificial intelligence, to identify salient variables and generate predictive models for forecasting trends and price changes in pharmaceutical claims data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If static data models are used for forecasting, then the system structure is simple and easy to implement, but the prediction accuracy deteriorates due to evolving trends in pharmaceutical data
Solution Approach 1:
The patent transforms static data models into dynamic predictive models that continuously adapt to evolving pharmaceutical data trends. The system uses machine learning algorithms to automatically update predictions based on new data, making the model dynamic rather than static, thereby resolving the contradiction between implementation simplicity and prediction accuracy.
Solution Approach 2:
The system changes the parameters of the forecasting model from fixed static values to dynamic learned parameters. By using training data to continuously adjust model parameters, the system maintains high prediction accuracy while the underlying model structure remains manageable through automated parameter optimization.
2Reliability
If manual verification steps are added to improve accuracy, then prediction reliability improves, but the productivity and processing speed deteriorate
Solution Approach 1:
The system implements self-service through automated machine learning models that perform verification and validation internally without requiring external manual intervention. The predictive models automatically assess their own confidence levels and can self-correct through continuous learning, maintaining high reliability while preserving processing speed.
Solution Approach 2:
The patent replaces manual verification processes with automated computational verification using machine learning algorithms. This substitution eliminates the need for human reviewers while maintaining or improving reliability through systematic automated validation, thereby preserving productivity.
3Measurement precision
If deep learning models with multiple algorithms are implemented, then prediction accuracy improves, but the device complexity increases
Solution Approach 1:
The patent segments the complex prediction task into distinct components handled by specialized algorithms: LSTM for temporal sequence modeling, MLP for non-linear pattern recognition, and gradient boosting for feature importance. This segmentation allows each algorithm to focus on specific aspects of the data, improving overall accuracy while making the system more manageable through modular architecture.
Solution Approach 2:
The system achieves multi-functionality by combining multiple algorithms that can handle different types of patterns in pharmaceutical data. The ensemble approach allows the system to universally handle various forecasting challenges (temporal trends, non-linear relationships, feature interactions) within a single integrated framework, justifying the increased complexity through enhanced versatility and accuracy.
Data Source
AI summary
A machine learning system is provided for training a data model to predict data states. The machine learning server is configured to receive a first portion of historical pharmaceutical data. The machine learning server is configured to apply a deep learning variable importance method to the first portion to identify at least one salient variable. The machine learning server is also configured to apply the model generation algorithm to the first portion and the at least one salient variable to generate predictive models for the forecast of the target variable. The machine learning server is also configured to receive a second portion of the historical pharmaceutical data to test the predictive models. The machine learning server is also configured to obtain a portion of current pharmaceutical data and apply the portion of current pharmaceutical data to the candidate predictive model to obtain the forecast of the target variable.