17993740. METHOD AND APPARATUS WITH NEURAL NETWORK simplified abstract (SAMSUNG ELECTRONICS CO., LTD.)
Contents
METHOD AND APPARATUS WITH NEURAL NETWORK
Organization Name
Inventor(s)
Junhaeng Lee of Hwaseong-si (KR)
Yeongjae Choi of Changwon-si (KR)
METHOD AND APPARATUS WITH NEURAL NETWORK - A simplified explanation of the abstract
This abstract first appeared for US patent application 17993740 titled 'METHOD AND APPARATUS WITH NEURAL NETWORK
Simplified Explanation
The abstract describes a method for training a neural network by updating the weights of its connections. The method involves calculating individual update values for each weight, generating an accumulated update value by adding these individual values, and updating the weight when the accumulated value reaches a threshold. The threshold is determined based on the significance of the bits in the weight.
- The method calculates individual update values for weights in a neural network.
- It generates an accumulated update value by adding these individual values.
- The weight is updated when the accumulated value reaches a threshold.
- The threshold is determined based on the significance of the bits in the weight.
Potential Applications
- This method can be applied in various fields where neural networks are used, such as image recognition, natural language processing, and autonomous vehicles.
- It can improve the efficiency and accuracy of training neural networks.
Problems Solved
- The method addresses the challenge of efficiently updating weights in a neural network during training.
- It ensures that weights are updated only when the accumulated update value reaches a specific threshold, reducing unnecessary updates.
Benefits
- The method improves the training process of neural networks by optimizing weight updates.
- It can lead to faster convergence and improved performance of neural networks.
- The threshold-based approach helps in avoiding unnecessary weight updates, saving computational resources.
Original Abstract Submitted
A processor-implemented neural network method includes calculating individual update values for a weight assigned to a connection relationship between nodes included in a neural network; generating an accumulated update value by adding the individual update values; and training the neural network by updating the weight using the accumulated update value in response to the accumulated update value being equal to or greater than a threshold value, wherein the threshold value is a value of 2 of an n-th bit of the weight, where the n-th bit is a bit of lesser significance than a bit in the weight representing a largest magnitude bit among all bits of the weight
