Adaptive Feature Selection for Mixed Data Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems struggle to effectively handle high-dimensional data with a large number of features compared to observations, particularly in mixed classification problems involving categorical, continuous, and functional data.
Innovation Solution
A computer-implemented method and system that performs adaptive feature selection by solving an optimizing function using an expanded dual augmented lagrangian algorithm, capable of handling mixed features and reducing dimensionality to optimize classification tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional classification systems are used on high-dimensional data, then the system can process the data, but the classification accuracy deteriorates due to the large number of features compared to observations
Solution Approach 1:
The patent extracts and selects only the most relevant features from the high-dimensional dataset using adaptive feature selection methods. This reduces the feature dimensionality while retaining the information necessary for accurate classification, directly addressing the problem of too many features relative to observations.
Solution Approach 2:
The patent changes the parameter of feature dimensionality by applying dimensionality reduction techniques. It transforms the original high-dimensional feature space into a lower-dimensional space that preserves the essential classification information, thereby improving classification accuracy when the number of features exceeds the number of observations.
2Reliability
If all features are used in classification, then comprehensive information is available, but computational complexity increases significantly
Solution Approach 1:
The patent extracts only the most informative features for classification, removing redundant and irrelevant features. This reduces the computational burden of processing all features while maintaining the reliability of classification performance by keeping the essential features.
Solution Approach 2:
The patent applies partial action by selecting a subset of features rather than processing all features. This partial feature selection reduces computational complexity while maintaining sufficient classification performance, avoiding the excessive computational cost of using all available features.
3Adaptability or versatility
If mixed feature types (categorical, continuous, functional) are handled separately, then each feature type can be processed appropriately, but the overall system complexity increases
Solution Approach 1:
The patent creates a universal feature selection framework that can handle multiple feature types (categorical, continuous, and functional) through a single integrated system. This multi-functional approach maintains adaptability to different feature types while reducing overall system complexity by avoiding separate processing pipelines for each feature type.
Solution Approach 2:
The patent segments the feature selection process into type-specific preprocessing steps followed by a unified selection mechanism. Each feature type undergoes appropriate preprocessing (e.g., one-hot encoding for categorical, smoothing for functional) before being fed into a common adaptive feature selection algorithm, balancing adaptability with system simplicity.
Data Source
AI summary
Embodiments receive a matrix of a plurality of observations and a plurality of features; Perform feature selection of the matrix of the plurality of observations and the plurality of features by solving an optimizing function; and output a class probability prediction based on estimated coefficient parameter values for selected features based on performing the feature selection of the matrix. In further embodiments, the plurality of features are mixed and include categorical features, functional features, and continuous features.