• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.65-82
• /
• 2021

Stock market investors are generally split into foreign investors, institutional investors, and individual investors. Compared to individual investor groups, professional investor groups such as foreign investors have an advantage in information and financial power and, as a result, foreign investors are known to show good investment performance among market participants. The purpose of this study is to propose an investment strategy that combines investor-specific transaction information and machine learning, and to analyze the portfolio investment performance of the proposed model using actual stock price and investor-specific transaction data. The Korea Exchange offers daily information on the volume of purchase and sale of each investor to securities firms. We developed a data collection program in C# programming language using an API provided by Daishin Securities Cybosplus, and collected 151 out of 200 KOSPI stocks with daily opening price, closing price and investor-specific net purchase data from January 2, 2007 to July 31, 2017. The self-organizing map model is an artificial neural network that performs clustering by unsupervised learning and has been introduced by Teuvo Kohonen since 1984. We implement competition among intra-surface artificial neurons, and all connections are non-recursive artificial neural networks that go from bottom to top. It can also be expanded to multiple layers, although many fault layers are commonly used. Linear functions are used by active functions of artificial nerve cells, and learning rules use Instar rules as well as general competitive learning. The core of the backpropagation model is the model that performs classification by supervised learning as an artificial neural network. We grouped and transformed investor-specific transaction volume data to learn backpropagation models through the self-organizing map model of artificial neural networks. As a result of the estimation of verification data through training, the portfolios were rebalanced monthly. For performance analysis, a passive portfolio was designated and the KOSPI 200 and KOSPI index returns for proxies on market returns were also obtained. Performance analysis was conducted using the equally-weighted portfolio return, compound interest rate, annual return, Maximum Draw Down, standard deviation, and Sharpe Ratio. Buy and hold returns of the top 10 market capitalization stocks are designated as a benchmark. Buy and hold strategy is the best strategy under the efficient market hypothesis. The prediction rate of learning data using backpropagation model was significantly high at 96.61%, while the prediction rate of verification data was also relatively high in the results of the 57.1% verification data. The performance evaluation of self-organizing map grouping can be determined as a result of a backpropagation model. This is because if the grouping results of the self-organizing map model had been poor, the learning results of the backpropagation model would have been poor. In this way, the performance assessment of machine learning is judged to be better learned than previous studies. Our portfolio doubled the return on the benchmark and performed better than the market returns on the KOSPI and KOSPI 200 indexes. In contrast to the benchmark, the MDD and standard deviation for portfolio risk indicators also showed better results. The Sharpe Ratio performed higher than benchmarks and stock market indexes. Through this, we presented the direction of portfolio composition program using machine learning and investor-specific transaction information and showed that it can be used to develop programs for real stock investment. The return is the result of monthly portfolio composition and asset rebalancing to the same proportion. Better outcomes are predicted when forming a monthly portfolio if the system is enforced by rebalancing the suggested stocks continuously without selling and re-buying it. Therefore, real transactions appear to be relevant.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.38 no.4
• /
• pp.375-382
• /
• 2020

As the 4th Industrial Revolution era in worldwide, interest in autonomous vehicles is increasing. but due to recent safety issues such as pedestrian accidents and car accidents, as a technical model for this, the demand for 3D HD maps (High Definition maps) is increasing in including lanes, road markings, road information, traffic lights and traffic signs etc. However, since some complementary points have been continuously raised according to demand, It is necessary to collect the opinions of institutions and companies utilizing HD maps and to improve HD maps. This study was conducted by utilizing the results of the contest for usability verification of HD Maps hosted by the National Geographic Information Institute and organized by the Spatial Information Industry Promotion Institute. For this study, we researched HD maps' layers and codes for HD maps object usability to improve HD maps, constructed HD maps object usability items accordingly, and contested usability verification of HD maps according to the items The contestants conducted verification and analyzed the results. As a result, the most frequently used code for each layer was the flat intersection, and the code showing the highest usage rate was a safety sign. In addition, the use rate of the sub-section and height obstacles was 16.67% and 8.88%, respectively, showing a low ratio. In order to utilize HD maps in the future, this study is expected to require research to continuously collect opinions from customers and improve data objects and data models that are actually needed by customers.

• Journal of the Korean earth science society
• /
• v.36 no.3
• /
• pp.280-299
• /
• 2015

The purpose of this study was to investigate the effects of an instruction using geologic planar figures on high school students' ability of spatial visualization and geologic spatial ability and also to explore its applicability as an instructional strategy through the investigation of students' perception about the instruction using mixed methodology. For this purpose, we developed 10 planar figures of geologic structures (2 horizontal layers, 2 vertical layers, 4 angular layers, 1 fault, and 1 fold), and tested students' spatial visualization ability and geologic spatial ability before and after the implementation in class. In addition, in order to investigate students' perception on the instruction, we conducted quantitative research using questionnaires comprised of the cognitive and the affective domain, and followed by focus group interview that was conducted to obtain deeper understanding of their perception. Findings revealed that the instruction using geologic planar figures was effective to enhance spatial visualization ability and geologic spatial ability. It was also helpful for students to enhance their ability to perceive the spatial configuration of the geologic structures as well as the ability to penetrate visually into the images of the structures. The results of the students' perception on the instruction showed that the students recognized the instruction using geologic planar figures as a strongly positive teaching method both in the cognitive and affective domain. We concluded that geologic planar figures could be used as an effective tool for the lesson of 'mapping and interpreting of geological map', and be highly applicable for the advanced class in high schools.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.32 no.5
• /
• pp.505-513
• /
• 2014

The RapidEye can acquire the 6.5m spatial resolution satellite imagery with the high temporal resolution on each day, based on its constellation of five satellites. The image products are available in two processing levels of Basic 1B and Ortho 3A. The Basic 1B image have radiometric and sensor corrections and include RPCs (Rational Polynomial Coefficients) data. In Korea, the geometric accuracy of RapidEye imagery can be improved, based on the scaled national digital maps that had been built. In this paper, we present the fully automated procedures to georegister the 1B data using 1:25,000 digital maps. Those layers of map are selected if the layers appear well in the RapidEye image, and then the selected layers are RPCs-projected into the RapidEye 1B space for generating vector images. The automated edge-based matching between the vector image and RapidEye improves the accuracy of RPCs. The experimental results showed the accuracy improvement from 2.8 to 0.8 pixels in RMSE when compared to the maps.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.3
• /
• pp.33-42
• /
• 2020

There has been difficulties in identifying objects by relying on the naked eye in various surveillance systems. There is a growing need for automated surveillance systems to replace soldiers in the field of military surveillance operations. Even though the object detection technology is developing rapidly in the civilian domain, but the research applied to the military is insufficient due to a lack of data and interest. Thus, in this paper, we applied one of deep learning algorithms, Convolutional Neural Network-based binary classification to develop an autonomous identification model of both friend and foe helicopters (AH-64, Mi-17) among the military weapon systems, and evaluated the model performance by considering accuracy, precision, recall and F-measure. As the result, the identification model demonstrates 97.8%, 97.3%, 98.5%, and 97.8 for accuracy, precision, recall and F-measure, respectively. In addition, we analyzed the feature map on convolution layers of the identification model in order to check which area of imagery is highly weighted. In general, rotary shaft of rotating wing, wheels, and air-intake on both of ally and foe helicopters played a major role in the performance of the identification model. This is the first study to attempt to classify images of helicopters among military weapons systems using CNN, and the model proposed in this study shows higher accuracy than the existing classification model for other weapons systems.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.6
• /
• pp.525-533
• /
• 2019

There are two feature collection methods in digital mapping using the UAV (Unmanned Aerial Vehicle) Photogrammetry: vectorization and stereo plotting. In vectorization, planar information is extracted from orthomosaics and elevation value obtained from a DSM (Digital Surface Model) or a DEM (Digital Elevation Model). However, the exact determination of the positional accuracy of 3D features such as ground facilities and buildings is very ambiguous, because the accuracy of vectorizing results has been mainly analyzed using only check points placed on the ground. Thus, this study aims to review the possibility of 3D spatial information acquisition and digital map production of vectorization by analyzing the corner point coordinates of different layers as well as check points. To this end, images were taken by a Phantom 4 (DJI) with 3.6 cm of GSD (Ground Sample Distance) at altitude of 90 m. The outcomes indicate that the horizontal RMSE (Root Mean Square Error) of vectorization method is 0.045 cm, which was calculated from residuals at check point compared with those of the field survey results. It is therefore possible to produce a digital topographic (plane) map of 1:1,000 scale using ortho images. On the other hand, the three-dimensional accuracy of vectorization was 0.068~0.162 m in horizontal and 0.090~1.840 m in vertical RMSE. It is thus difficult to obtain 3D spatial information and 1:1,000 digital map production by using vectorization due to a large error in elevation.

• Journal of Intelligence and Information Systems
• /
• v.15 no.2
• /
• pp.49-59
• /
• 2009

Contents-based image retrieval methods are in general more objective and effective than text-based image retrieval algorithms since they use color and texture in search and avoid annotating all images for search. SIM(Self-organizing Image browsing Map) is one of contents-based image retrieval algorithms that uses only browsable mapping results obtained by SOM(Self Organizing Map). However, SOM may have an error in selecting the right BMU in learning phase if there are similar nodes with distorted color information due to the intensity of light or objects' movements in the image. Such images may be mapped into other grouping nodes thus the search rate could be decreased by this effect. In this paper, we propose an improved SIM that uses HSV color model in extracting image features with color quantization. In order to avoid unexpected learning error mentioned above, our SOM consists of two layers. In learning phase, SOM layer 1 has the color feature vectors as input. After learning SOM Layer 1, the connection weights of this layer become the input of SOM Layer 2 and re-learning occurs. With this multi-layered SOM learning, we can avoid mapping errors among similar nodes of different color information. In search, we put the query image vector into SOM layer 2 and select nodes of SOM layer 1 that connects with chosen BMU of SOM layer 2. In experiment, we verified that the proposed SIM was better than the original SIM and avoid mapping error effectively.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.36 no.4
• /
• pp.221-232
• /
• 2023

This paper presents a novel convolutional neural network (CNN)-based approach for predicting bolt clamping force in the early bolt loosening state of bolted structures. The approach entails tightening eight bolts with different clamping forces and generating frequency responses, which are then used to create a similarity map. This map quantifies the magnitude and shape similarity between the frequency responses and the initial model in a fully fastened state. Krylov subspace-based model order reduction is employed to efficiently handle the large amount of frequency response data. The CNN model incorporates a regression output layer to predict the clamping forces of the bolts. Its performance is evaluated by training the network by using various amounts of training data and convolutional layers. The input data for the model are derived from the magnitude and shape similarity map obtained from the frequency responses. The results demonstrate the diagnostic potential and effectiveness of the proposed approach in detecting early bolt loosening. Accurate bolt clamping force predictions in the early loosening state can thus be achieved by utilizing the frequency response data and CNN model. The findings afford valuable insights into the application of CNNs for assessing the integrity of bolted structures.

• Journal of Soil and Groundwater Environment
• /
• v.28 no.6
• /
• pp.71-89
• /
• 2023

A groundwater potential map (GPM) was built for the Nakdonggang River Basin based on ten variables, including hydrogeologic unit, fault-line density, depth to groundwater, distance to surface water, lineament density, slope, stream drainage density, soil drainage, land cover, and annual rainfall. To integrate the thematic layers for GPM, the criteria were first weighted using the Analytic Hierarchical Process (AHP) and then overlaid using the Technique for Ordering Preferences by Similarity to Ideal Solution (TOPSIS) model. Finally, the groundwater potential was categorized into five classes (very high (VH), high (H), moderate (M), low (L), very low (VL)) and verified by examining the specific capacity of individual wells on each class. The wells in the area categorized as VH showed the highest median specific capacity (5.2 m³/day/m), while the wells with specific capacity < 1.39 m³/day/m were distributed in the areas categorized as L or VL. The accuracy of GPM generated in the work looked acceptable, although the specific capacity data were not enough to verify GPM in the studied large watershed. To create GPMs for the determination of high-yield well locations, the resolution and reliability of thematic maps should be improved. Criterion values for groundwater potential should be established when machine learning or statistical models are used in the GPM evaluation process.

• Journal of Korean Society for Atmospheric Environment
• /
• v.13 no.6
• /
• pp.463-474
• /
• 1997

Metallic elements in dustfall have been known as notable air pollutants directly or indirectly influencing human health and wealth. The first aim of this study was to obtain precise spatial distribution patterns of 5 elements (Pb, Zn, K, Cr, and Al) in dustfall around Suwon area. To predict isometric lines of metal fluxes deposited on unsupervised random sites, the study has applied both spatial statistics as a receptor model and a GIS (geographic information system). Total of 31 sampling sites were selected in the study area (roughly 3 by 3 km grid basis) and dustfall samples were then collected monthly basis by the British deposit gauges from Dec., 1995 to Nov., 1996. The metallic elements in the dustfall were then analyzed by an atomic absorption spectrometer (AAS). On the other hand, a base map overlapped by 7 layers was constructed by using the AutoCAD R13 and ARC/INFO 3.4D. Four different spatial interpolation and expolation techniques such as IDW (inverse distance weighted averaging), TIN (triangulated irregular network), polynomial regression, and kriging technique were examined to compare spatial distribution patterns. Each pattern obtained by each technique was substantally different as varing pollutant types, land of use types, and topological conditions, etc. Thus, our study focused intensively on uncertainty analysis based on a concept of the jackknife and the sum of error distance. It was found that a kriging technique was the best applicalbe in this study area.