17550551. BOOSTING CLASSIFICATION AND REGRESSION TREE PERFORMANCE WITH DIMENSION REDUCTION simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)

From WikiPatents
Jump to navigation Jump to search

BOOSTING CLASSIFICATION AND REGRESSION TREE PERFORMANCE WITH DIMENSION REDUCTION

Organization Name

INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor(s)

Dzung Tien Phan of Pleasantville NY (US)

Michael Huang of Los Angeles CA (US)

Pavankumar Murali of Ardsley NY (US)

Lam Minh Nguyen of Ossining NY (US)

BOOSTING CLASSIFICATION AND REGRESSION TREE PERFORMANCE WITH DIMENSION REDUCTION - A simplified explanation of the abstract

This abstract first appeared for US patent application 17550551 titled 'BOOSTING CLASSIFICATION AND REGRESSION TREE PERFORMANCE WITH DIMENSION REDUCTION

Simplified Explanation

The patent application describes a system and method for constructing and training a decision tree for machine learning. Here is a simplified explanation of the abstract:

  • The system receives a training set for machine learning.
  • It initializes a decision tree by creating a root node.
  • A root solver is trained using the training set.
  • The decision tree is grown by splitting nodes iteratively.
  • At each node, dimension reduction is performed on the features of the training data.
  • The data with reduced dimension is split based on a routing function to another node.
  • The dimension reduction and split are performed together at the node using a nonlinear optimization problem.

Potential applications of this technology:

  • Machine learning: The decision tree can be used for classification or regression tasks in various domains such as finance, healthcare, and marketing.
  • Data analysis: The dimension reduction technique can help in analyzing high-dimensional data and extracting meaningful insights.
  • Pattern recognition: The decision tree can be used to identify patterns and make predictions based on the trained model.

Problems solved by this technology:

  • Efficient decision tree construction: The system provides a method to construct a decision tree by iteratively splitting nodes, reducing the complexity of the process.
  • Dimension reduction: The system performs dimension reduction on the training data, which helps in reducing noise and improving the accuracy of the decision tree.
  • Nonlinear optimization: The system solves a nonlinear optimization problem to perform dimension reduction and split data, allowing for more flexible and accurate decision tree construction.

Benefits of this technology:

  • Improved accuracy: By performing dimension reduction and splitting data based on a routing function, the decision tree can be trained to make more accurate predictions.
  • Faster training: The iterative splitting of nodes and the use of nonlinear optimization help in speeding up the training process of the decision tree.
  • Flexibility: The system allows for the construction of decision trees that can handle complex and nonlinear relationships in the data, making it suitable for a wide range of applications.


Original Abstract Submitted

A system and method can be provided for constructing and training a decision tree for machine learning. A training set can be received. The decision tree can be initialized by constructing a root node and a root solver can be trained with the training set. A processor can grow the decision tree by iteratively splitting nodes of the decision tree, where at a node of the decision tree, dimension reduction is performed on features of data of the training set received at the node, and the data having reduced dimension is split based on a routing function, for routing to another node of the decision tree. The dimension reduction and the split can be performed together at the node based on solving a nonlinear optimization problem.

Retrieved from "http://wikipatents.org/index.php?title=17550551._BOOSTING_CLASSIFICATION_AND_REGRESSION_TREE_PERFORMANCE_WITH_DIMENSION_REDUCTION_simplified_abstract_(INTERNATIONAL_BUSINESS_MACHINES_CORPORATION)&oldid=26255"