18239542. KERNEL LEARNING APPARATUS USING TRANSFORMED CONVEX OPTIMIZATION PROBLEM simplified abstract (NEC Corporation)
KERNEL LEARNING APPARATUS USING TRANSFORMED CONVEX OPTIMIZATION PROBLEM
Organization Name
Inventor(s)
KERNEL LEARNING APPARATUS USING TRANSFORMED CONVEX OPTIMIZATION PROBLEM - A simplified explanation of the abstract
This abstract first appeared for US patent application 18239542 titled 'KERNEL LEARNING APPARATUS USING TRANSFORMED CONVEX OPTIMIZATION PROBLEM
Simplified Explanation
The patent application describes a kernel learning apparatus that preprocesses and represents data examples using feature representations. It then uses an explicit feature mapping circuit to design a kernel function that embeds the feature representations into a nonlinear feature space. This allows for training a predictive model using the explicit feature map. The apparatus also includes a convex problem formulating circuitry that formulates a non-convex problem into a convex optimization problem based on the explicit feature map. Finally, an optimal solution solving circuitry solves the convex optimization problem to obtain a globally optimal solution for training an interpretable predictive model.
- Data preprocessing circuitry preprocesses and represents data examples as feature representations.
- Explicit feature mapping circuit designs a kernel function with an explicit feature map to embed the feature representations into a nonlinear feature space.
- Convex problem formulating circuitry formulates a non-convex problem into a convex optimization problem based on the explicit feature map.
- Optimal solution solving circuitry solves the convex optimization problem to obtain a globally optimal solution for training an interpretable predictive model.
Potential Applications
- Machine learning and predictive modeling
- Natural language processing
- Image and video analysis
- Financial forecasting
Problems Solved
- Nonlinear feature representation and embedding
- Training interpretable predictive models
- Formulating non-convex problems into convex optimization problems
Benefits
- Improved accuracy and interpretability of predictive models
- Efficient training of models in nonlinear feature spaces
- Ability to solve non-convex problems using convex optimization techniques
Original Abstract Submitted
In a kernel learning apparatus, a data preprocessing circuitry preprocesses and represents each data example as a collection of feature representations that need to be interpreted. An explicit feature mapping circuit designs a kernel function with an explicit feature map to embed the feature representations of data into a nonlinear feature space and to produce the explicit feature map for the designed kernel function to train a predictive model. A convex problem formulating circuitry formulates a non-convex problem for training the predictive model into a convex optimization problem based on the explicit feature map. An optimal solution solving circuitry solves the convex optimization problem to obtain a globally optimal solution for training an interpretable predictive model.
