18333554. MACHINE LEARNING METHOD, INFORMATION PROCESSING SYSTEM, INFORMATION PROCESSING APPARATUS, SERVER, AND PROGRAM simplified abstract (FUJIFILM Corporation)
MACHINE LEARNING METHOD, INFORMATION PROCESSING SYSTEM, INFORMATION PROCESSING APPARATUS, SERVER, AND PROGRAM
Organization Name
Inventor(s)
Tomoki Taniguchi of Tokyo (JP)
MACHINE LEARNING METHOD, INFORMATION PROCESSING SYSTEM, INFORMATION PROCESSING APPARATUS, SERVER, AND PROGRAM - A simplified explanation of the abstract
This abstract first appeared for US patent application 18333554 titled 'MACHINE LEARNING METHOD, INFORMATION PROCESSING SYSTEM, INFORMATION PROCESSING APPARATUS, SERVER, AND PROGRAM
Simplified Explanation
The patent application describes an information processing system that uses data from multiple facilities to train a local model that predicts user behavior on an item at each facility. The system then evaluates the difference between the models and corrects the training to minimize the difference.
- The system includes one or more processors that perform the training of a local model using data from each facility.
- The local model predicts user behavior on an item at each facility.
- The system evaluates the difference between the models in a parameter of the local model trained for each facility.
- Based on the evaluation, the system corrects the training of the local model to minimize the difference between the models.
Potential applications of this technology:
- E-commerce platforms can use this system to predict user behavior on different items at different facilities, allowing for personalized recommendations.
- Online advertising platforms can utilize this system to predict user behavior on ads at different facilities, enabling targeted advertising campaigns.
- Social media platforms can employ this system to predict user behavior on posts or content at different facilities, improving content recommendations.
Problems solved by this technology:
- The system solves the problem of predicting user behavior on items at different facilities by training a local model for each facility.
- It addresses the challenge of evaluating and minimizing the difference between models to ensure accurate predictions across facilities.
Benefits of this technology:
- The system allows for personalized predictions and recommendations based on user behavior at different facilities.
- It improves the accuracy of predictions by evaluating and correcting the training of the local models.
- The technology enables targeted advertising and content recommendations, leading to better user engagement and satisfaction.
Original Abstract Submitted
An information processing system includes one or more processors, in which the one or more processors are configured to: perform a training, by using data of each facility collected at each of a plurality of facilities, of a local model that predicts, for each of the facilities, a behavior of a user on an item; perform an evaluation of a difference between models in a parameter of the local model trained for each of the facilities; and correct the training of the local model such that the difference between the models is small based on a result of the evaluation.
