17525932. COMPOSITE FEATURE ENGINEERING simplified abstract (International Business Machines Corporation)

COMPOSITE FEATURE ENGINEERING

Organization Name

International Business Machines Corporation

Inventor(s)

Dakuo Wang of Cambridge MA (US)

Udayan Khurana of White Plains NY (US)

Chuang Gan of Cambridge MA (US)

Gregory Bramble of Larchmont NY (US)

Abel Valente of Buenos Aires (AR)

Arunima Chaudhary of Dehradun (IN)

Carolina Maria Spina of Olavarria (AR)

Micah Smith of Boston MA (US)

COMPOSITE FEATURE ENGINEERING - A simplified explanation of the abstract

This abstract first appeared for US patent application 17525932 titled 'COMPOSITE FEATURE ENGINEERING

Simplified Explanation

The patent application describes a method for identifying domain knowledge features in a dataset and generating predictive models based on user-defined features. Here are the key points:

• The method starts by identifying the domain of an input dataset.
• Archived domain knowledge features related to the identified domain are then identified.
• User-defined feature definitions are inputted by the user.
• The identified archived domain knowledge features and user-defined features are processed together.
• This processing generates a set of candidate features that can be presented to the user.
• The user selects a subset of the candidate features.
• Based on the selected features, one or more predictive models are generated.

Potential Applications

This technology has potential applications in various fields where predictive modeling is used, such as:

• Financial analysis and forecasting
• Healthcare diagnostics and prediction
• Customer behavior analysis and prediction
• Fraud detection and prevention
• Supply chain optimization

Problems Solved

This technology addresses the following problems:

• Identifying relevant domain knowledge features in a dataset can be time-consuming and challenging.
• User-defined features may not always capture all the necessary information for accurate predictive modeling.
• Selecting the most relevant features from a large set of candidates can be overwhelming for users.
• Generating predictive models based on selected features can be complex and resource-intensive.

Benefits

The use of this technology offers several benefits:

• Efficient identification of relevant domain knowledge features saves time and effort.
• Incorporating user-defined features allows for customization and domain-specific modeling.
• The generation of predictive models based on selected features improves accuracy and efficiency.
• The method can be applied to various datasets and domains, increasing its versatility.

Original Abstract Submitted

A domain of an input dataset is identified and one or more archived domain knowledge features corresponding to the identified domain are identified. One or more user feature definitions for one or more user features defined by a user are inputted. The identified archived domain knowledge features and the user features are processed to generate a set of candidate features for presentation to the user. A selection of a subset of the candidate features is obtained from the user and one or more predictive models are generated based on the selected features.