Neural Network Dummy Layer for Reverse Decision Variable Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine learning models, particularly those used in cell culture processes, cannot deduce decision variables in reverse to achieve specific results, leading to labor-intensive and resource-consuming design and improvement processes.
Innovation Solution
A method is introduced that involves adding a dummy layer to a pre-trained neural network predictive model, with artificial neurons connected to the input layer, setting bias values to 0, and using an optimizer to adjust weight values to match target results, allowing for the calculation of decision variables.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional machine learning predictive models are used to predict cell culture outcomes, then model accuracy and convergence speed are improved, but the ability to deduce decision variables in reverse is lost, resulting in labor-intensive and resource-consuming process design
Solution Approach 1:
The patent inverts the traditional forward prediction approach by implementing a reverse deduction mechanism. Instead of inputting decision variables to predict outcomes, the system inputs target outcomes and deduces the corresponding decision variables that would produce those outcomes. This is achieved by adding a dummy layer to the neural network and using the optimizer to solve for input parameters that minimize the difference between predicted and target results, thereby enabling reverse engineering of optimal process parameters.
2Adaptability or versatility
If multiple decision variables are manually adjusted to achieve desired cell culture results, then treatment customization is improved, but the complexity and resource consumption of the manufacturing process increases
Solution Approach 1:
The patent implements self-service by enabling the system to automatically determine optimal decision variables without requiring manual adjustment. The neural network model, equipped with the dummy layer and optimizer, autonomously calculates the input parameters needed to achieve target outcomes. This self-determining capability maintains high adaptability for customized treatments while eliminating the manual complexity associated with adjusting multiple decision variables for each cell preparation.
Data Source
AI summary
The present invention provides a method for calculating decision variables. A dummy layer is added at an input layer of a trained neural network predictive model. The dummy layer includes a plurality of artificial neurons respectively connected to a corresponding input terminal of the input layer for the trained predictive model by a newly established link. The input value of each artificial neuron is set to 1, the bias value of the activation function is set to 0, and the output of the activation function is set to 1 when the input of the activation function is 1. The initial weight value of the newly established link is selected and set, and the weight values can be considered as decision variables, wherein the weight values can have ranges or other inter-conditional restrictions. The optimal solution is obtained using the optimizer built in a general machine-learning platform when the parameters of the trained predictive model are frozen, and only the weight values of the newly established links are adjusted. The training objective is set so that the output of the parameter predictive model matches the desired target result. At the end of the training, the weight values of the newly established link new are the feasible input decision variables. This invention allows the effective use of the general machine learning platforms and the built-in methods to find the optimal input parameters that achieve the expected results.