Lg Innotek Co., Ltd.

Samsung Electronics Co., Ltd.

Lg Electronics Inc.

Korea Electronics Technology Institute

Riiid Inc.

Korea University Research and Business Foundation

Goldstar, Co., Ltd.

Other

Seoul National University

Yonsei University

Dongbu Daewoo Electronics Corporation

Byung Soo Kim

Keun Sik Lee

Hyuk Jin Hong

Chung Won Seo

Ji Won Jo

Sun Hwa Lee

Kyoung Joung Kim

Jung Eun Kim

Yong In Lee

Min Su Song

Young Kyu Kim

Young Duck Choi

Young Jong Jang

Yong In Park

Il Han Park

Sang-Soo Park

Jong-Hoon Lee

Kyoung Hoon Chai

Jae We Heo

Jong Il Kim

Tae Ho Hwang

Jung Hyun Cho

Jae-Yun Lee

Yeong Min Cha

Dong Min Shin

Young Nam Lee

Ji-Hye Jung

Joon Soo Kwon

Chul Min Lim

Jin Eon Baek

See Woo Lee

Jae Min Lee

Kyung Tae Kim

Kyung Mi Park

Kang-Bin Lee

Yoon Heung Tak

Chan Kyu Koo

Na Ri Lee

Jung Han Ryu

Su-Yong Kim

Hong Koo Baik

Hyun Moon Park

Tae Il Lee

Seung Ho Kwon

Hyun Min Kim

Jung Sung Kim

Jin Hwan Lee

Hyun Bin Loh

June Young Park

Woo Soon Jang

Chan Soo Shin

Na-Young Choi

Jae Yong Ahn

Yun Mi Park

Ja Kyo Koo

Chang Hun Lee

Hee Tak Kim

Chan Bae

Suk Ho Bhang

Wan Geun La

Jeong A Shin

Jin Ho Sunwoo

Chang Hun Jun

Jung Soo Han

Ki Sik Min

Pil Jae Chae

Seo Yon Park

Seok Hoon Jeon

Provided is an apparatus for training a knowledge tracking model, which is an apparatus for predicting a correct answer probability of a user on the basis of data augmentation, the apparatus including: a problem-solving data storage unit configured to store problem-solving data in which a problem solved by a user and a response of the user to the problem are mapped; a data augmentation performing unit configured to receive the problem-solving data from the problem-solving storage unit and convert the problem-solving data to generate augmented data; a regularization performing unit configured to receive the augmented data from the data augmentation performing unit and perform a regularization operation using a regularization loss function determined on the basis of a data augmentation method that is performed; and a model training unit configured to input the augmented data to a knowledge tracking model, allow the knowledge tracking model to learn a weight representing a relationship between a problem-solving record included in the augmented data and a correct answer probability of the user, and train the knowledge tracking model for predicting a correct answer probability of the user for a specific problem on the basis of the weight.

Apparatus and system for training knowledge tracking model based on data augmentation and operation method thereof

The present disclosure relates to a system capable of increasing the accuracy of a prediction result when predicting a user drop-out rate or predicting a correct answer probability of a user in an online learning environment, and further relates to an operation method of the system. A system according to the present disclosure may include a plurality of encoder neural networks and a plurality of decoders, and may input question information to a kencoder neural network and input response information to kdecoder neural network to learn the system, thereby predicting the user drop-out rate information and the users correct answer probability information with higher accuracy on the basis of the learned system.

System for predicting user drop-out rate and tracking user knowledge based on artificial intelligence learning and method therefor

According to an embodiment of the present disclosure, a pre-training modeling system for predicting an educational element using a pre-trained AI model includes: a pre-training unit configured to pre-train an AI model, through an evaluation element that can be automatically collected every time a user solves an individual question by receiving an interaction element according to an offline user's operation from a user terminal, classifying an AI element of educational content into the interaction element and the educational element, and intersecting the interaction element and the educational element; a fine-tuning unit configured to perform an additional training operation with respect to the pre-trained AI model, according to an educational element to be predicted; and an educational element prediction unit configured to predict, on the basis of the fine-tuned AI model, one or more pieces of information among test scores, grades, qualifications, and review information indicating whether the user got a correct answer when re-solving the same question, which are information that must be separately input from the user because the amount of information automatically collected every time the user solves an individual question is small or cannot be automatically collected.

Pre-training modeling system and method for predicting educational element

Provided are a hyperparameter optimizer and method for optimizing hyperparameters and a spiking neural network processing unit. The optimizer includes a statistical analyzer configured to receive training data and perform statistical analysis on the training data, an objective function generator configured to generate hyperparameter-specific objective functions by using a statistical analysis value of the statistical analyzer, and an optimal hyperparameter selector configured to select optimal hyperparameters according to certain rules on the basis of the hyperparameter-specific objective functions.

Hyper-parameter optimization method for spiking neural network and the processing apparatus thereof

The present disclosure relates to a method for classifying instructions according to the number of operands required for processing-in-memory instruction processing, and a computing device applying same. Efficient instruction processing in a processing-in-memory may include identifying the number of operands required when processing an instruction queuing to be processed, interpreting the instruction queuing to be processed and processing an instruction corresponding to the identified number of required operands. When the number of required operands is 0, the instruction interpretation may interpret the instruction queuing to be processed as a WRITE instruction, and the instruction processing may execute memory writing. When the number of required operands is not 0, the instruction processing may execute memory reading in an internal memory of the processing-in-memory by the same number of times as the number of operands required in the instruction interpreted in the instruction interpretation.

Processing-in-memory control method for efficient instruction processing and computing device applying same

This application relates to a frictional electricity generator increasing output transfer efficiency. In one aspect, the frictional electricity generator includes a housing and a power generation element which is arranged in the housing and in which a + position pole and a &#x2212; position pole are arranged according to the movement of a positive charge and a negative charge by friction. The + position pole and the &#x2212; position pole of the power generation element may be arranged at a distance of a predetermined angle on a plane so as to minimize energy loss due to mutual repulsion, and may be connected to each transmission path.

Frictional electricity generator increasing output transfer efficiency

This application relates to a method and apparatus for processing a new read-write-operation instruction added to an instruction set to maximize the performance of processing-in-memory (PIM). The read-write-operation instruction performs reading and writing on an operation result of the PIM by returning the operation result of the PIM to a computer system and, at the same time, writing the operation result to a destination address. An instruction processor in PIM includes a response data selector and a finite state machine to process the read-write-operation instruction. The response data selector includes a selector configured to select one of a response data signal and an operation result, and a three-phase buffer configured to allow or disallow response data. The finite state machine of the instruction processor outputs a response permission signal and a response selection signal for controlling the buffer and the selector.

Method and apparatus for processing read-write-operation instruction in processing-in-memory

This application relates to a memory management method for maximizing processing-in-memory (PIM) performance and reducing unnecessary DRAM access time. In one aspect, when processing a PIM instruction packet, an instruction processing unit secondarily processes a request for access to a destination address at which read and write actions of an internal memory are likely to be sequentially performed. By secondarily requesting the destination address, a row address of an open page of the internal memory may match a row address to which a PIM instruction packet processing result is written back. Also, the instruction processing unit inside the PIM maintains memory write and read addresses that have previously requested. The instruction processing unit compares the address of a packet to be processed to the maintained previous memory address and informs a memory controller about the comparison result through a page closing signal.

Memory management method and apparatus for processing-in-memory

A memory device comprises a memory cell array including a first memory cell disposed on a substrate and a second memory cell above the first memory cell; a first word line connected to the first memory cell and a second word line connected to the second memory cell, the second word line disposed above the first word line; and a word line defect detection circuit configured to monitor a number of pulses of a pumping clock signal while applying a first voltage to the first word line to detect a defect of the first word line. The voltage generator is configured to apply a second voltage different from the first voltage to the second word line for programming the second memory cell when the number of pulses of the pumping clock signal is smaller than a reference value.

Memory device including a circuit for detecting word line defect and operating method thereof

The present invention relates to a method of inducing the differentiation of human pluripotent stem cell-derived embryoid bodies into ectoderm by CXCR2 stimulation, and more particularly, to a method of promoting the differentiation of human pluripotent stem cell-derived embryoid bodies into ectoderm by stimulating and activating the surface receptor CXCR2 of the embryoid bodies with the CXCR2-specific ligand GRO-&#x3b1;. The method of inducing the differentiation of human pluripotent stem cell-derived embryoid bodies into ectoderm by CXCR2 stimulation according to the present invention can increase the efficiency and utility of stem cells as a cell therapeutic agent, because it promotes the differentiation of stem cells into a specific germ layer serving as the origin of target cells, which is the first important step for inducing the differentiation of stem cells into specific cells.

Method for inducing ectodermal differentiation of embryoid bodies derived from human pluripotent stem cells by CXCR2 stimulation

Disclosed are a touch panel and a method for manufacturing the same. A touch panel can include a substrate, a transparent electrode base on the substrate, a first transparent electrode on the transparent electrode base and extending in a first direction, and a second transparent electrode on the transparent electrode base and extending in a second direction. A method of manufacturing a touch panel can include preparing a substrate and a transparent electrode base, forming a transparent electrode over the transparent electrode base, and forming an electrode material over the transparent electrode base.

Touch panel and method for manufacturing the same

A memory device can include: a memory cell array including a memory cell and a word line that is connected to the memory cell; a clock generator configured to generate a first pumping clock signal from a system clock signal; a charge pump configured to provide a pumping voltage signal using a power supply voltage and the first pumping clock signal; a compensation circuit configured to compensate for variations in a first reference clock signal in accordance with variations in the power supply voltage, and provide a compensated first reference clock signal; and a pass/fail (P/F) determining circuit configured to determine whether the word line is defective by comparing the first pumping clock signal and the compensated first reference clock signal while the pumping voltage signal is provided to the word line.

Memory devices including a word line defect detection circuit

A medical patch having an electric potential generating structure which is attachable to a wound to thus regenerate injured skin tissues is provided. The medical patch includes a unit patch having a piezoelectric potential generating structure. The unit patch includes a first layer, a second layer and a piezoelectric nanomaterial disposed between the first and second layers.

Medical patch

Disclosed are a touch panel and a method for manufacturing the same. The touch panel according to one embodiment includes the touch panel including a substrate, a first transparent electrode including a first sensor part formed on the substrate in a first direction and a first electrode connection part for electric connection of the first sensor part, and a second transparent electrode electrically insulated from the first transparent electrode, and including a second sensor part formed in a second direction crossing the first direction and a second electrode connection part for electric connection of the second sensor part. The second electrode connection part has a mesh shape. The touch panel according to another embodiment includes a substrate, a transparent electrode base spaced apart from the substrate, and a transparent electrode on the transparent electrode base. The transparent electrode comprises a first electrode including a first sensor part formed in one direction and a first electrode connection part for electrical connection of the first sensor part; and a second electrode, which is electrically insulated from the first electrode, and includes a second sensor part formed in a second direction crossing the first direction and a second electrode connection part for electrical connection of the second sensor part. The second electrode connection part has a mesh shape.

Growing community of inventors

Seoul, South Korea

Byung Soo Kim