• Animal cells and systems
• /
• v.14 no.1
• /
• pp.11-15
• /
• 2010

Long-term working memory (LTWM) is a subdivision concept of working memory and indicates the enhancement of performance in a working memory task. LTWM has been shown in humans who have been engaged in a specific task requiring working memory over a long time. However, there is very little understanding of the exact mechanism of LTWM because of limitations of experimental methods in human studies. We have modified the standard T-maze task, which is used to test working memory in mice, to demonstrate LTWM in an animal model. We observed an enhancement of performance by repeated experience with the same working memory load in mice, which can be regarded as an LTWM. This effect seems to depend on the condition wherein a delay was given. This task may be a good experimental protocol to assess LTWM in animal studies.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.7
• /
• pp.920-926
• /
• 2004

The human being receives a new information from outside and the information shows gradual oblivion with time. But the information remains in memory and isn't forgotten for a long time if the information is read several times over. For example, we assume that we memorize a telephone number when we listen and never remind we may forget it soon, but we commit to memory long time by repeating. If the human being received new information with strong stimulus, it could remain in memory without recalling repeatedly. The moments of almost losing one's life in an accident or getting a stroke of luck are rarely forgiven. The human being can keep memory for a long time in spite of the limit of memory for the mechanism mentioned above. In this paper, we propose a model to explain the mechanism mentioned above using a neural network and fuzzy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.3
• /
• pp.837-852
• /
• 2021

Recurrent neural network (RNN) architectures have been used for language modeling (LM) tasks that require learning long-range word or character sequences. However, the RNN architecture is still suffered from unstable gradients on long-range sequences. To address the issue of long-range sequences, an attention mechanism has been used, showing state-of-the-art (SOTA) performance in all LM tasks. A differentiable neural computer (DNC) is a deep learning architecture using an attention mechanism. The DNC architecture is a neural network augmented with a content-addressable external memory. However, in the write operation, some information unrelated to the input word remains in memory. Moreover, DNCs have been found to perform poorly with low numbers of weight parameters. Therefore, we propose a robust memory deallocation method using a limited retention vector. The limited retention vector determines whether the network increases or decreases its usage of information in external memory according to a threshold. We experimentally evaluate the robustness of a DNC implementing the proposed approach according to the size of the controller and external memory on the enwik8 LM task. When we decreased the number of weight parameters by 32.47%, the proposed DNC showed a low bits-per-character (BPC) degradation of 4.30%, demonstrating the effectiveness of our approach in language modeling tasks.

• Korean Journal of Biological Psychiatry
• /
• v.9 no.2
• /
• pp.79-94
• /
• 2002

Polarisation of biological and psychosocial aspects of psychiatry is nowadays main stream. Current knowledges of the interaction between biology and psychology make it possible to consider a truly integrative approach of the two aspects. Research findings suggest that the neuronal plasticity is the key mechanism to answer how the mental function work to an environmental stimuli and how the psychotherapeutic approach work on the brain. Advances in neuroscience research have led to a more sophisticated understanding of how psychotherapy may affect brain function. Even though there have been a tremendous efforts to find out the neurobiological mechanism of mental function, the answer is at best premature. In this article, research findings about of neuronal plasticity, implicit memory, animal studies which were associated with psychotherapy and psychological aspects were reviewed.

• Journal of the Semiconductor & Display Technology
• /
• v.5 no.1 s.14
• /
• pp.17-20
• /
• 2006

This paper presents of a new type of memory cell that could potentially replace both DRAM and flash memory. The proposed device cell operates by sensing the state of about 1,000 electrons trapped between unique insulating barriers in the channel region of the upper transistor. These electrons are controlled by a side gate on the transistor, and their state in turn controls the gate of the larger transistor, providing signal gain within the memory cell. It becomes faster and more reliable memory with lower operation voltage. Moreover, the use of a multiple tunnel junction (MTJ) fur the vertical transistor can significantly improve the data retention and operation speed.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.21 no.7
• /
• pp.615-619
• /
• 2008

A molecular memory device which has a structure of Al/$Al_2O_3$/ASA-15 LB monolayer/Ti/Al device, was fabricated. To study a charge transfer mechanism of molecular memory devices, current density-voltage (J-V) characteristics were measured at an increasing temperature range from 10 K to 300 K with an interval of 30 K. Strong temperature-dependent electrical property and tunneling through organic monolayer at low bias (below 0.5 V) were appeared. These experimental data were fitted by using a theoretical formula such as the Simmons model. In comparison between the theoretical and the experimental results, it was verified that the fitting results using the Simmons model about direct tunneling was fairly fitted below 0.5 V at both 300 K and 10 K. Hopping conduction was also dominant at all voltage range above 200 K due to charges trapped by defects located within the dielectric stack, including the $Al_2O_3$, organic monolayer and Ti interfaces.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.17 no.5
• /
• pp.70-75
• /
• 2008

This paper describes a tool-clamping/unclamping mechanism for application of a micro-spindle. The mechanism is based on one-way shape memory effect and interference-fit. The corresponding mathematical models and a few considerable design parameters are mentioned in this paper. Especially, necessary conditions for the clamping and unclamping operation are investigated through finite element analysis. The analysis results show that the differences between the diametral deformations of the tool holder in high temperature and that in low temperature are increased according to amounts of the interference. Thus the less interference between the tool-holder and the ring, the less tolerance to allow the clamping and unclamping operation because the inner diameter of the tool holder in high temperature should be smaller than the diameter of the tool shank, and that in low temperature should be larger than the diameter of the tool shank. In addition, the design for maximization of clamping force are investigated based on finite element analysis. The results show that the more amounts of the interference, the more clamping force. As the result, the interference should be considered as a important factor to maximize the tool clamping force.

• Journal of Korea Society of Industrial Information Systems
• /
• v.26 no.1
• /
• pp.1-10
• /
• 2021

This study proposed a Long-Short Term Memory network to consider changes in emotion over time, and applied an attention mechanism to give weights to the emotion states that appear at specific moments. We used 32 channel EEG data from DEAP database. A 2-level classification (Low and High) experiment and a 3-level classification experiment (Low, Middle, and High) were performed on Valence and Arousal emotion model. As a result, accuracy of the 2-level classification experiment was 90.1% for Valence and 88.1% for Arousal. The accuracy of 3-level classification was 83.5% for Valence and 82.5% for Arousal.

• Korean Journal of Cognitive Science
• /
• v.30 no.4
• /
• pp.239-268
• /
• 2019

Since last several decades, iconic memory has been accepted theoretically valid for its role of the first storage mechanism in visual memory process. However, there have been relatively less interests in iconic memory among researchers than their interests in visual short- and long-term memory. Such little interests seem to arise from a lack of detailed understandings of theories and methodologies about iconic memory and visual persistence. This study aimed to achieve the understandings by reviewing theories and empirical studies of iconic memory and visual persistence. The study further discussed future direction of iconic memory research according to the essential aspects of phenomenological and ecological criticisms against the validity of iconic memory.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.38 no.10
• /
• pp.24-33
• /
• 2001

Delay, delay variation and packet loss rate are principal QoS(Quality of Service) elements of packet communication. This paper proposes a new buffer allocation mechanism to improve the packet loss performance in such a situation that multiple logical buffers share a single physical memory buffer. In the proposed buffer allocation mechanism, the movement of dynamic threshold follows a curved track instead of a straight line which is used in the DT(dynamic threshold) mechanism. In order evaluate the effectiveness of the proposed mechanism, it is compared with the existing previously proposed mechanisms in several aspects including NC(no control), ST(Static Threshold) and DT mechanisms.