Machine Learning Prediction Using Precomputed Vector Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting conditions using machine learning models are inefficient due to the selection of irrelevant features and lack of robust feature determination, leading to inaccurate likelihood scores.
Innovation Solution
A machine learning model is developed using pre-computed vectors derived from a training dataset, with feature selection based on statistical significance and random forest variable importance, ensuring only relevant features with non-zero coefficients are included, and utilizing algorithms like random forests for prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pre-computed vectors are derived from training dataset to preprocess target data, then normalization accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-computing vectors from the training dataset before actual prediction tasks. These pre-computed vectors are stored and reused for normalizing target data, eliminating the need to recompute normalization parameters for each new prediction, thus improving normalization accuracy while reducing computational complexity during deployment.
2Reliability
If feature selection is performed using statistical significance and random forest variable importance, then model reliability is improved, but processing time increases
Solution Approach 1:
The patent performs feature selection using statistical significance tests and random forest variable importance analysis during the model training phase. This preliminary action identifies and selects the most relevant features beforehand, ensuring model reliability. The selected features are then used for all subsequent predictions, avoiding repeated feature selection processing and reducing time loss during actual deployment.
3Measurement precision
If irrelevant features are removed from the dataset, then prediction accuracy is improved, but information loss may occur
Solution Approach 1:
The patent extracts and removes irrelevant features from the dataset using statistical significance analysis and random forest variable importance. By identifying features that do not contribute meaningfully to prediction (e.g., those with p-values above threshold or low variable importance scores), the system extracts only the relevant features for modeling. This selective extraction improves prediction accuracy by eliminating noise while preserving essential information contained in relevant features.
Data Source
AI summary
In some embodiments, a machine learning model may be accessed and used to generate a likelihood score related to a condition. In some embodiments, pre-computed vectors may be derived from a training dataset used to build the machine learning model, and the pre-computed vectors may be used to generate processed data from target data derived from a target sample. The machine learning model may then be used on the processed data to generate the likelihood score related to the condition. As an example, subsets of the training dataset may be randomly selected, and the pre-computed vectors may be derived from the randomly-selected subsets of the training dataset. The pre-computed vectors may be applied to the target data to generate the processed data. In one use case, for example, the target data may be normalized using the pre-computed vectors.