International business machines corporation (20240127084). JOINT PREDICTION AND IMPROVEMENT FOR MACHINE LEARNING MODELS simplified abstract

JOINT PREDICTION AND IMPROVEMENT FOR MACHINE LEARNING MODELS

Organization Name

international business machines corporation

Inventor(s)

Yuya Jeremy Ong of San Jose CA (US)

Aly Megahed of San Jose CA (US)

Mark S. Squillante of Greenwich CT (US)

Yingdong Lu of Yorktown Heights NY (US)

Yitao Liang of Sherman Oaks CA (US)

Pravar Mahajan of Sunnyvale CA (US)

JOINT PREDICTION AND IMPROVEMENT FOR MACHINE LEARNING MODELS - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240127084 titled 'JOINT PREDICTION AND IMPROVEMENT FOR MACHINE LEARNING MODELS

Simplified Explanation

The abstract describes a method for a joint prediction and improvement framework for machine learning models. Here is a simplified explanation of the abstract:

• Obtaining a machine learning model with a set of parameters
• Identifying actions based on test inputs and historical actions related to a task
• Using the identified actions to compute values for inference loss and computing cost function
• Updating the parameters of the machine learning model based on the computed values

---

1. 1. 1. Potential Applications of this Technology

- Enhancing the performance of machine learning models - Improving the accuracy of predictions in various tasks

1. 1. 1. Problems Solved by this Technology

- Addressing the need for continuous improvement of machine learning models - Streamlining the process of updating model parameters based on historical actions

1. 1. 1. Benefits of this Technology

- Increased efficiency in machine learning model optimization - Enhanced predictive capabilities in various applications

1. 1. 1. Potential Commercial Applications of this Technology
         1. Optimizing Machine Learning Models for Enhanced Performance

1. 1. 1. Possible Prior Art

- Prior methods for updating machine learning model parameters based on historical actions and test inputs

---

1. 1. 1. Unanswered Questions

1. 1. 1. 1. How does this method compare to existing techniques for model optimization?

The article does not provide a direct comparison to existing methods for model optimization. Further research or experimentation may be needed to evaluate the effectiveness of this approach in comparison to other techniques.

1. 1. 1. 1. What are the potential limitations or challenges of implementing this joint prediction and improvement framework?

The article does not discuss potential limitations or challenges that may arise when implementing this framework. Understanding these factors could be crucial for successful adoption and integration of the method into existing machine learning workflows.

Original Abstract Submitted

methods, systems, and computer program products for a joint prediction and improvement framework for machine learning models are provided herein. a method includes obtaining a machine learning model initialized with a set of parameters; identifying one or more actions based on test inputs corresponding to the machine learning model and historical actions related to a task, where the historical actions are dependent on respective historical outputs of the machine learning model; using the identified one or more actions to jointly compute: one or more first values corresponding to inference loss for the machine learning model; and one or more second values based at least in part on a computing cost function associated with the task; and updating the set of parameters of the machine learning model based on the one or more first values and the one or more second values.