• Dementia and Neurocognitive Disorders
• /
• v.21 no.1
• /
• pp.30-41
• /
• 2022

Background and Purpose: In this study we aimed to find the association between neuropsychological performance and body mass index (BMI) in patients with mild cognitive impairment (MCI). In addition, we investigated the effects of the apolipoprotein E (APOE) genotype in the relationship between the BMI and cognition in MCI. Methods: We enrolled a cohort of 3,038 subjects with MCI aged 65-90 from the Clinical Research Center for Dementia of South Korea and a dementia cohort of the Ewha Womans University Mokdong Hospital. MCI patients were classified into three subgroups according to the Asian standard of BMI. We compared cognitive performances between groups by one-way analysis of variance. To investigate the effects of the APOE genotype, we used multivariate linear regression models after adjusting for possible confounders. Results: Even though normal BMI groups were younger, had more females, and had less comorbidities, the higher BMI groups had better cognitive functions. Among subjects with APOE ε4 carriers, there was a positive relationship between the BMI and the memory task alone. Conclusions: Our findings suggested that higher BMI in patients with MCI were associated with better cognitive performance. The effects of the APOE ε4 genotype in the associations between BMI and cognition were distinguishing. Therefore, according to physical status, APOE ε4 genotype-specific strategies in the assessments and treatments may be necessary in elderly patients with MCI.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.9
• /
• pp.351-360
• /
• 2018

The LWR (Locally Weighted Regression) model, which is traditionally a lazy learning model, is designed to obtain the solution of the prediction according to the input variable, the query point, and it is a kind of the regression equation in the short interval obtained as a result of the learning that gives a higher weight value closer to the query point. We study on an incremental ensemble learning approach for LWR, a form of lazy learning and memory-based learning. The proposed incremental ensemble learning method of LWR is to sequentially generate and integrate LWR models over time using a genetic algorithm to obtain a solution of a specific query point. The weaknesses of existing LWR models are that multiple LWR models can be generated based on the indicator function and data sample selection, and the quality of the predictions can also vary depending on this model. However, no research has been conducted to solve the problem of selection or combination of multiple LWR models. In this study, after generating the initial LWR model according to the indicator function and the sample data set, we iterate evolution learning process to obtain the proper indicator function and assess the LWR models applied to the other sample data sets to overcome the data set bias. We adopt Eager learning method to generate and store LWR model gradually when data is generated for all sections. In order to obtain a prediction solution at a specific point in time, an LWR model is generated based on newly generated data within a predetermined interval and then combined with existing LWR models in a section using a genetic algorithm. The proposed method shows better results than the method of selecting multiple LWR models using the simple average method. The results of this study are compared with the predicted results using multiple regression analysis by applying the real data such as the amount of traffic per hour in a specific area and hourly sales of a resting place of the highway, etc.

• KIPS Transactions on Computer and Communication Systems
• /
• v.7 no.6
• /
• pp.155-164
• /
• 2018

Many malicious programs have been compressed or encrypted using various commercial packers to prevent reverse engineering, So malicious code analysts must decompress or decrypt them first. The OEP (Original Entry Point) is the address of the first instruction executed after returning the encrypted or compressed executable file back to the original binary state. Several unpackers, including PinDemonium, execute the packed file and keep tracks of the addresses until the OEP appears and find the OEP among the addresses. However, instead of finding exact one OEP, unpackers provide a relatively large set of OEP candidates and sometimes OEP is missing among candidates. In other words, existing unpackers have difficulty in finding the correct OEP. We have developed new tool which provides fewer OEP candidate sets by adding two methods based on the property of the OEP. In this paper, we propose two methods to provide fewer OEP candidate sets by using the property that the function call sequence and parameters are same between packed program and original program. First way is based on a function call. Programs written in the C/C++ language are compiled to translate languages into binary code. Compiler-specific system functions are added to the compiled program. After examining these functions, we have added a method that we suggest to PinDemonium to detect the unpacking work by matching the patterns of system functions that are called in packed programs and unpacked programs. Second way is based on parameters. The parameters include not only the user-entered inputs, but also the system inputs. We have added a method that we suggest to PinDemonium to find the OEP using the system parameters of a particular function in stack memory. OEP detection experiments were performed on sample programs packed by 16 commercial packers. We can reduce the OEP candidate by more than 40% on average compared to PinDemonium except 2 commercial packers which are can not be executed due to the anti-debugging technique.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.71-88
• /
• 2017

Language models were originally developed for speech recognition and language processing. Using a set of example sentences, a language model predicts the next word or character based on sequential input data. N-gram models have been widely used but this model cannot model the correlation between the input units efficiently since it is a probabilistic model which are based on the frequency of each unit in the training set. Recently, as the deep learning algorithm has been developed, a recurrent neural network (RNN) model and a long short-term memory (LSTM) model have been widely used for the neural language model (Ahn, 2016; Kim et al., 2016; Lee et al., 2016). These models can reflect dependency between the objects that are entered sequentially into the model (Gers and Schmidhuber, 2001; Mikolov et al., 2010; Sundermeyer et al., 2012). In order to learning the neural language model, texts need to be decomposed into words or morphemes. Since, however, a training set of sentences includes a huge number of words or morphemes in general, the size of dictionary is very large and so it increases model complexity. In addition, word-level or morpheme-level models are able to generate vocabularies only which are contained in the training set. Furthermore, with highly morphological languages such as Turkish, Hungarian, Russian, Finnish or Korean, morpheme analyzers have more chance to cause errors in decomposition process (Lankinen et al., 2016). Therefore, this paper proposes a phoneme-level language model for Korean language based on LSTM models. A phoneme such as a vowel or a consonant is the smallest unit that comprises Korean texts. We construct the language model using three or four LSTM layers. Each model was trained using Stochastic Gradient Algorithm and more advanced optimization algorithms such as Adagrad, RMSprop, Adadelta, Adam, Adamax, and Nadam. Simulation study was done with Old Testament texts using a deep learning package Keras based the Theano. After pre-processing the texts, the dataset included 74 of unique characters including vowels, consonants, and punctuation marks. Then we constructed an input vector with 20 consecutive characters and an output with a following 21st character. Finally, total 1,023,411 sets of input-output vectors were included in the dataset and we divided them into training, validation, testsets with proportion 70:15:15. All the simulation were conducted on a system equipped with an Intel Xeon CPU (16 cores) and a NVIDIA GeForce GTX 1080 GPU. We compared the loss function evaluated for the validation set, the perplexity evaluated for the test set, and the time to be taken for training each model. As a result, all the optimization algorithms but the stochastic gradient algorithm showed similar validation loss and perplexity, which are clearly superior to those of the stochastic gradient algorithm. The stochastic gradient algorithm took the longest time to be trained for both 3- and 4-LSTM models. On average, the 4-LSTM layer model took 69% longer training time than the 3-LSTM layer model. However, the validation loss and perplexity were not improved significantly or became even worse for specific conditions. On the other hand, when comparing the automatically generated sentences, the 4-LSTM layer model tended to generate the sentences which are closer to the natural language than the 3-LSTM model. Although there were slight differences in the completeness of the generated sentences between the models, the sentence generation performance was quite satisfactory in any simulation conditions: they generated only legitimate Korean letters and the use of postposition and the conjugation of verbs were almost perfect in the sense of grammar. The results of this study are expected to be widely used for the processing of Korean language in the field of language processing and speech recognition, which are the basis of artificial intelligence systems.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• Investigative Magnetic Resonance Imaging
• /
• v.9 no.1
• /
• pp.43-49
• /
• 2005

A few studies have shown that the function of fusiform face area is selectively involved in the perception of faces including a race difference. We investigated the neural substrates of the face-selective region called fusiform face area in the ventral occipital-temporal cortex and same-race memory superiority in the fusiform face area by the event-related fMRI. In our fMRI study, subjects (Oriental-Korean) performed the implicit distinction of the race while they consciously made familiar-judgments, regardless of whether they considered a face as Oriental-Korean or European-American. For race distinction as an implicit task, the fusiform face areas (FFA) and the right parahippocampal gyrus had a greater response to the presentation of Oriental-Korean faces than for the European-American faces, but in the conscious race distinction between Oriental-Korean and European-American faces, there was no significant difference observed in the FFA. These results suggest that different activation in the fusiform regions and right parahippocampal gyrus resulting from superiority of same-race memory could have implicitly taken place by the physiological processes of face recognition.

• Spatial Information Research
• /
• v.20 no.5
• /
• pp.99-109
• /
• 2012

Recently, due to the growth of social media and spread of smart-phone, the amount of data has considerably increased by full use of SNS (Social Network Service). According to it, the Big Data concept is come up and many researchers are seeking solutions to make the best use of big data. To maximize the creative value of the big data held by many companies, it is required to combine them with existing data. The physical and theoretical storage structures of data sources are so different that a system which can integrate and manage them is needed. In order to process big data, MapReduce is developed as a system which has advantages over processing data fast by distributed processing. However, it is difficult to construct and store a system for all key words. Due to the process of storage and search, it is to some extent difficult to do real-time processing. And it makes extra expenses to process complex event without structure of processing different data. In order to solve this problem, the existing Complex Event Processing System is supposed to be used. When it comes to complex event processing system, it gets data from different sources and combines them with each other to make it possible to do complex event processing that is useful for real-time processing specially in stream data. Nevertheless, unstructured data based on text of SNS and internet articles is managed as text type and there is a need to compare strings every time the query processing should be done. And it results in poor performance. Therefore, we try to make it possible to manage unstructured data and do query process fast in complex event processing system. And we extend the data complex function for giving theoretical schema of string. It is completed by changing the string key word into integer type with filtering which uses keyword set. In addition, by using the Complex Event Processing System and processing stream data at real-time of in-memory, we try to reduce the time of reading the query processing after it is stored in the disk.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1993.06a
• /
• pp.975-976
• /
• 1993

This talk presents the overview of the author's research and development activities on fuzzy inference hardware. We involved it with two distinct approaches. The first approach is to use application specific integrated circuits (ASIC) technology. The fuzzy inference method is directly implemented in silicon. The second approach, which is in its preliminary stage, is to use more conventional microprocessor architecture. Here, we use a quantitative technique used by designer of reduced instruction set computer (RISC) to modify an architecture of a microprocessor. In the ASIC approach, we implemented the most widely used fuzzy inference mechanism directly on silicon. The mechanism is beaded on a max-min compositional rule of inference, and Mandami's method of fuzzy implication. The two VLSI fuzzy inference chips are designed, fabricated, and fully tested. Both used a full-custom CMOS technology. The second and more claborate chip was designed at the University of North Carolina(U C) in cooperation with MCNC. Both VLSI chips had muliple datapaths for rule digital fuzzy inference chips had multiple datapaths for rule evaluation, and they executed multiple fuzzy if-then rules in parallel. The AT & T chip is the first digital fuzzy inference chip in the world. It ran with a 20 MHz clock cycle and achieved an approximately 80.000 Fuzzy Logical inferences Per Second (FLIPS). It stored and executed 16 fuzzy if-then rules. Since it was designed as a proof of concept prototype chip, it had minimal amount of peripheral logic for system integration. UNC/MCNC chip consists of 688,131 transistors of which 476,160 are used for RAM memory. It ran with a 10 MHz clock cycle. The chip has a 3-staged pipeline and initiates a computation of new inference every 64 cycle. This chip achieved an approximately 160,000 FLIPS. The new architecture have the following important improvements from the AT & T chip: Programmable rule set memory (RAM). On-chip fuzzification operation by a table lookup method. On-chip defuzzification operation by a centroid method. Reconfigurable architecture for processing two rule formats. RAM/datapath redundancy for higher yield It can store and execute 51 if-then rule of the following format: IF A and B and C and D Then Do E, and Then Do F. With this format, the chip takes four inputs and produces two outputs. By software reconfiguration, it can store and execute 102 if-then rules of the following simpler format using the same datapath: IF A and B Then Do E. With this format the chip takes two inputs and produces one outputs. We have built two VME-bus board systems based on this chip for Oak Ridge National Laboratory (ORNL). The board is now installed in a robot at ORNL. Researchers uses this board for experiment in autonomous robot navigation. The Fuzzy Logic system board places the Fuzzy chip into a VMEbus environment. High level C language functions hide the operational details of the board from the applications programme . The programmer treats rule memories and fuzzification function memories as local structures passed as parameters to the C functions. ASIC fuzzy inference hardware is extremely fast, but they are limited in generality. Many aspects of the design are limited or fixed. We have proposed to designing a are limited or fixed. We have proposed to designing a fuzzy information processor as an application specific processor using a quantitative approach. The quantitative approach was developed by RISC designers. In effect, we are interested in evaluating the effectiveness of a specialized RISC processor for fuzzy information processing. As the first step, we measured the possible speed-up of a fuzzy inference program based on if-then rules by an introduction of specialized instructions, i.e., min and max instructions. The minimum and maximum operations are heavily used in fuzzy logic applications as fuzzy intersection and union. We performed measurements using a MIPS R3000 as a base micropro essor. The initial result is encouraging. We can achieve as high as a 2.5 increase in inference speed if the R3000 had min and max instructions. Also, they are useful for speeding up other fuzzy operations such as bounded product and bounded sum. The embedded processor's main task is to control some device or process. It usually runs a single or a embedded processer to create an embedded processor for fuzzy control is very effective. Table I shows the measured speed of the inference by a MIPS R3000 microprocessor, a fictitious MIPS R3000 microprocessor with min and max instructions, and a UNC/MCNC ASIC fuzzy inference chip. The software that used on microprocessors is a simulator of the ASIC chip. The first row is the computation time in seconds of 6000 inferences using 51 rules where each fuzzy set is represented by an array of 64 elements. The second row is the time required to perform a single inference. The last row is the fuzzy logical inferences per second (FLIPS) measured for ach device. There is a large gap in run time between the ASIC and software approaches even if we resort to a specialized fuzzy microprocessor. As for design time and cost, these two approaches represent two extremes. An ASIC approach is extremely expensive. It is, therefore, an important research topic to design a specialized computing architecture for fuzzy applications that falls between these two extremes both in run time and design time/cost. TABLEI INFERENCE TIME BY 51 RULES {{{{Time }}{{MIPS R3000 }}{{ASIC }}{{Regular }}{{With min/mix }}{{6000 inference 1 inference FLIPS }}{{125s 20.8ms 48 }}{{49s 8.2ms 122 }}{{0.0038s 6.4㎲ 156,250 }} }}

• Journal of the Korean Institute of Landscape Architecture
• /
• v.46 no.4
• /
• pp.21-35
• /
• 2018

In recent years, there has been a tendency to reuse 'landscapes of memory,' including industrial heritages, modern cultural heritages, and post-industrial parks, as public spaces in many cities. Among the various types of landscapes, 'modern historic landscapes', which were formed in the 19th and 20th centuries, are landscapes where the debate between conservation and removal is most frequent, according to the change of evaluation and recognition of modern history. This study examines conflicts between conservation and removal around modern historic landscapes and explores the value judgment criteria and the process of formation of those landscapes, as highlighted in the case of the demolition controversy of the old Japanese general government building in Seoul, which was dismantled in 1995. First, this study reviews newspaper articles, television news and debate programs from 1980-1999 and some articles related to the controversy of the Japanese general government building. Then it draws the following six factors as the main issues of the demolition controversy of the building: symbolic location, discoveries and responses of new historical facts, reaction and intervention of a related country, financial conditions, function and usage of the landscape, changes of urban, historical and architectural policies. Based on these issues, this study examines the conflicts between symbolic values that play an important role in the formation of modern historic landscapes and determines conservation or removal, and the utility of functional values that solve the problems and respond to criticisms that arise in the process of forming the modern historic landscape. Especially, it is noted that the most important factor that makes the decision is the symbolic values, although the determination of the conservation or removal of modern historic landscapes has changed according to changes in historical perceptions of modern history. Today, the modern historic landscape is an important site for urban design, and still has historical issues to be agreed upon and addressed. Thi study has contemporary significance from the point that it divides the many values of modern historic landscapes into symbolic values and functional values, evaluates these, and reviews the background social context.

• Journal of Life Science
• /
• v.29 no.6
• /
• pp.747-753
• /
• 2019

Cognitive decline is characterized by reduced long-/short-term memory and attention span, and increased depression and anxiety. Such decline is associated with various degenerative brain disorders, especially Alzheimer's disease (AD) and Parkinson's disease (PD). The increases in elderly populations suffering from cognitive decline create social problems and impose economic burdens, and also pose safety threats; all of these problems have been extensively researched over the past several decades. Possible causes of cognitive decline include metabolic and hormone imbalance, infection, medication abuse, and neuronal changes associated with aging. However, no treatment for cognitive decline is available. In neurodegenerative diseases, changes in the gut microbiota and gut metabolites can alter molecular expression and neurobehavioral symptoms. Changes in the gut microbiota affect memory loss in AD via the downregulation of NMDA receptor expression and increased glutamate levels. Furthermore, the use of probiotics resulted in neurological improvement in an AD model. PD and gut microbiota dysbiosis are linked directly. This interrelationship affected the development of constipation, a secondary symptom in PD. In a PD model, the administration of probiotics prevented neuron death by increasing butyrate levels. Dysfunction of the blood-brain barrier (BBB) has been identified in AD and PD. Increased BBB permeability is also associated with gut microbiota dysbiosis, which led to the destruction of microtubules via systemic inflammation. Notably, metabolites of the gut microbiota may trigger either the development or attenuation of neurodegenerative disease. Here, we discuss the correlation between cognitive decline and the gut microbiota.