18237035. FEATURE INTERACTION USING ATTENTION-BASED FEATURE SELECTION simplified abstract (Micron Technology, Inc.)

FEATURE INTERACTION USING ATTENTION-BASED FEATURE SELECTION

Organization Name

Micron Technology, Inc.

Inventor(s)

Mritunjay Kumar of Bhagalpur (IN)

Tejashri Kelhe of Pune (IN)

Nidhi Nika of Sitamarhi (IN)

FEATURE INTERACTION USING ATTENTION-BASED FEATURE SELECTION - A simplified explanation of the abstract

This abstract first appeared for US patent application 18237035 titled 'FEATURE INTERACTION USING ATTENTION-BASED FEATURE SELECTION

Simplified Explanation

The system described in the patent application involves using attention-based feature selection and feature interaction to generate predictions from tabular data.

• The system obtains a set of base features associated with tabular data.
• It selects a set of relevant features from the base features using attention-based feature selection.
• The set of relevant features is a subset of the base features.
• It generates interaction features from the set of relevant features using feature interaction.
• Finally, the system generates predictions using the set of interaction features.

---

1. 1. 1. Potential Applications

• Predictive analytics in various industries such as finance, healthcare, and marketing.
• Personalized recommendations in e-commerce platforms.
• Fraud detection in financial transactions.

1. 1. 1. Problems Solved

• Efficient feature selection from large datasets.
• Improved prediction accuracy by considering feature interactions.
• Automation of prediction generation process.

1. 1. 1. Benefits

• Enhanced predictive modeling capabilities.
• Increased efficiency in data analysis tasks.
• Better decision-making based on accurate predictions.

Original Abstract Submitted

A system includes a memory and a processing device, operatively coupled to the memory, to perform operations including obtaining a set of base features associated with tabular data, selecting, from the set of base features, a set of relevant features using attention-based feature selection, wherein the set of relevant features is a subset of the set of base features, generating, from the set of relevant features using feature interaction, a set of interaction features, and generating a prediction using the set of interaction features.