17457698. ACCELERATING DECISION TREE INFERENCES BASED ON COMPLEMENTARY TENSOR OPERATION SETS simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)
ACCELERATING DECISION TREE INFERENCES BASED ON COMPLEMENTARY TENSOR OPERATION SETS
Organization Name
INTERNATIONAL BUSINESS MACHINES CORPORATION
Inventor(s)
Nikolaos Papandreou of Thalwil (CH)
Charalampos Pozidis of Thalwil (CH)
Milos Stanisavljevic of Adliswil (CH)
Jan Van Lunteren of Rüschlikon (CH)
Cedric Lichtenau of Stuttgart (DE)
Andrew M. Sica of Oxford CT (US)
ACCELERATING DECISION TREE INFERENCES BASED ON COMPLEMENTARY TENSOR OPERATION SETS - A simplified explanation of the abstract
This abstract first appeared for US patent application 17457698 titled 'ACCELERATING DECISION TREE INFERENCES BASED ON COMPLEMENTARY TENSOR OPERATION SETS
Simplified Explanation
The patent application describes a method for representing machine learning inferences using tensors.
- Tensors are used to represent subsets of decision trees based on statistics and data attributes.
- The subsets are ranked based on their corresponding leaf node subsets.
- This method allows for a more efficient representation of machine learning inferences.
Potential Applications
- This technology can be applied in various fields where machine learning is used, such as healthcare, finance, and marketing.
- It can be used for tasks like classification, regression, and anomaly detection.
Problems Solved
- Traditional methods of representing machine learning inferences can be computationally expensive and inefficient.
- This technology solves the problem of efficiently representing and ranking subsets of decision trees.
Benefits
- The use of tensors allows for a more compact and efficient representation of machine learning inferences.
- This method can improve the performance and speed of machine learning algorithms.
- It provides a more structured and organized way to analyze and interpret machine learning results.
Original Abstract Submitted
A tensor representation of a machine learning inferences to be performed is built by forming complementary tensor subsets that respectively correspond to complementary subsets of one or more leaf nodes of one or more decision trees based on statistics of the one or more leaf nodes of the one or more decision trees and data capturing attributes of one or more split nodes of the one or more decision trees and the one or more leaf nodes of the decision trees. The complementary tensor subsets are ranked such that a first tensor subset and a second tensor subset of the complementary tensor subsets correspond to a first leaf node subset and a second leaf node subset of the complementary subsets of the one or more leaf nodes.
