18094027. SEGMENT MODELING FOR MACHINE LEARNING USING TENSOR TRAIN DECOMPOSITIONS simplified abstract (MICROSOFT TECHNOLOGY LICENSING, LLC)

SEGMENT MODELING FOR MACHINE LEARNING USING TENSOR TRAIN DECOMPOSITIONS

Organization Name

MICROSOFT TECHNOLOGY LICENSING, LLC

Inventor(s)

Ajith Muralidharan of Sunnyvale CA (US)

Ankan Saha of San Francisco CA (US)

Prakruthi Prabhakar of Foster City CA (US)

SEGMENT MODELING FOR MACHINE LEARNING USING TENSOR TRAIN DECOMPOSITIONS - A simplified explanation of the abstract

This abstract first appeared for US patent application 18094027 titled 'SEGMENT MODELING FOR MACHINE LEARNING USING TENSOR TRAIN DECOMPOSITIONS

• • Simplified Explanation:**

The patent application discusses the use of tensor train decompositions to create personalized layers for efficient segment modeling. Instead of training on a matrix containing all segments, the data is mapped to an N-dimensional tensor, allowing for efficient training and scoring through tensor train decomposition.

• • Key Features and Innovation:**

- Utilizing tensor train decompositions for segment modeling - Mapping data to an N-dimensional tensor for efficient training - Creating personalized layers for improved scoring

• • Potential Applications:**

- Machine learning - Data analysis - Pattern recognition

• • Problems Solved:**

- Efficient training on large datasets - Personalized layer creation for segment modeling

• • Benefits:**

- Improved efficiency in training and scoring - Enhanced accuracy in segment modeling

• • Commercial Applications:**

Title: "Efficient Segment Modeling Technology for Enhanced Data Analysis" This technology can be applied in various industries such as finance, healthcare, and marketing for improved data analysis, pattern recognition, and machine learning applications. The market implications include increased efficiency, accuracy, and scalability in data processing tasks.

• • Prior Art:**

Readers can start searching for prior art related to tensor train decompositions, segment modeling, and efficient data analysis techniques in academic journals, patent databases, and relevant industry publications.

• • Frequently Updated Research:**

Stay updated on advancements in tensor train decompositions, segment modeling techniques, and applications in machine learning for enhanced data analysis.

• • Questions about Tensor Train Decompositions:**

1. How do tensor train decompositions differ from other matrix decomposition techniques?

  - Tensor train decompositions are unique in their ability to efficiently handle high-dimensional data by decomposing tensors into a series of smaller tensors.

2. What are the main advantages of using tensor train decompositions in segment modeling?

  - Tensor train decompositions offer improved efficiency, scalability, and accuracy in segment modeling tasks compared to traditional methods.

Original Abstract Submitted

In an example embodiment, tensor train decompositions are used to create large, personalized layers that are efficient for segment modeling. More particularly, rather than performing learning on an input matrix of training data that contains all segments, and then crossing this matrix with a vector for a particular segment, the matrix is mapped to an N-dimensional tensor, where each of the dimensions corresponds to one of the properties used to compose the segment, which can then be approximated by tensor train decomposition to enable efficient training and scoring.