• Journal of the Korea Convergence Society
• /
• v.12 no.4
• /
• pp.249-258
• /
• 2021

Bitcoin prices have been soaring recently as investors flock to cryptocurrency exchanges. The purpose of this study is to predict the Bitcoin price using a deep learning model and analyze whether Bitcoin is profitable through investment strategy. LSTM is utilized as Bitcoin prediction model with nonlinearity and long-term memory and the profitability of MA cross-over strategy with predicted prices as input variables is analyzed. Investment performance of Bitcoin strategy using LSTM forecast prices from 2013 to 2021 showed return improvement of 5.5% and 46% more than market price MA cross-over strategy and benchmark Buy & Hold strategy, respectively. The results of this study, which expanded to recent data, supported the inefficiency of the cryptocurrency market, as did previous studies, and showed the feasibility of using the deep learning model for Bitcoin investors. In future research, it is necessary to develop optimal prediction models and improve the profitability of Bitcoin investment strategies through performance comparison of various deep learning models.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.46 no.5
• /
• pp.1-13
• /
• 2009

Filter cache has been introduced as one solution of reducing cache power consumption. More than 50% of the power reduction results from the filter cache, whereas more than 20% of the performance is compromised. To minimize the performance degradation of the filter cache, the predictive filter cache has been proposed. In this paper, we review the previous filter cache predictors and analyze the problems of the solutions. As a result, we found main problems that cause prediction misses in previous filter cache schemes and, to resolve the problems, this paper proposes a new prediction policy. In our scheme, some reference bit entries, called MSBs, are inserted into filter cache and BTB, to adaptively control the filter cache access. In simulation parts, we use a modified SimpleScalar simulator with MiBench benchmark programs to verify the proposed filter cache. The simulation result shows in average 5% performance improvement, compared to previous ones.

• The KIPS Transactions:PartC
• /
• v.12C no.1 s.97
• /
• pp.69-76
• /
• 2005

To effectively use limited resources in wireless cellular networks it is necessary to predict exactly the amount of resources required by handoff calls at a future time. In this paper we propose a method which predicts the amount of resources needed by handoff calls more accurately than the existing method based on Wiener processes. The existing method uses the current demands to predict future demands. Although this method is much simpler than using traffic information from neighbor cells, its prediction error increases as time elapses, leading to waste of wireless resources. By using an exponential parameter to decrease the magnitude of error over time, we show in simulation how to outperform the existing method in resource utilization as well as in prediction of resource demands.

• KIPS Transactions on Computer and Communication Systems
• /
• v.4 no.11
• /
• pp.369-382
• /
• 2015

Energy aware server clusters aim to reduce power consumption at maximum while keeping QoS(quality of service) compared to energy non-aware server clusters. They adjust the power mode of each server in a fixed or variable time interval to activate only the minimum number of servers needed to handle current user requests. Previous studies on energy aware server cluster put efforts to reduce power consumption or heat dissipation, but they do not consider energy efficiency well. In this paper, we propose an energy efficient cluster management method to improve not only performance per watt but also QoS of the existing server power mode control method based on autonomous learning. Our proposed method is to adjust server power mode based on a hybrid approach of autonomous learning method with multi level thresholds and power consumption prediction method. Autonomous learning method with multi level thresholds is applied under normal load situation whereas power consumption prediction method is applied under abnormal load situation. The decision on whether current load is normal or abnormal depends on the ratio of the number of current user requests over the average number of user requests during recent past few minutes. Also, a dynamic shutdown method is additionally applied to shorten the time delay to make servers off. We performed experiments with a cluster of 16 servers using three different kinds of load patterns. The multi-threshold based learning method with prediction and dynamic shutdown shows the best result in terms of normalized QoS and performance per watt (valid responses). For banking load pattern, real load pattern, and virtual load pattern, the numbers of good response per watt in the proposed method increase by 1.66%, 2.9% and 3.84%, respectively, whereas QoS in the proposed method increase by 0.45%, 1.33% and 8.82%, respectively, compared to those in the existing autonomous learning method with single level threshold.

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.4
• /
• pp.2229-2233
• /
• 2016

Background: Breast cancer is one of the most common cancers among women in the world. Early detection is necessary to improve outcomes and decrease related costs. The aim of this study was to assess the predictive power of health locus of control as a modifying factor in the Health Belief Model (HBM) for prediction of breast self-examination. Materials and Methods: In this cross- sectional study, 400 women selected through the convenience sampling from health centers. Data were collected using part of the Champion's HBM scale (CHBMS), the Health Locus of Control Scale and a self administered questionnaire. For data analysis by SPSS the independent T test, Chi square test, logistic and linear regression modes were appliedl. Results: The results showed that 10.9% of the participants reported performing BSE regularly. Health locus of control did not act as a predictor of BSE as a modifying factor. In this study, perceived self-efficacy was the strongest predictor of BSE performance (Exp (B) =1.863) with direct effect, while awareness had direct and indirect influence. Conclusions: For increasing BSE, improvement of self-efficacy especially in young women and increasing knowledge about cancer is necessary.

• Advances in concrete construction
• /
• v.13 no.1
• /
• pp.11-23
• /
• 2022

Conventional concrete needs some improvement in the mechanical properties, which can be obtained by different admixtures. However, making concrete samples costume always time and money. In this paper, different types of hybrid algorithms are applied to develop predictive models for forecasting compressive strength (CS) of concretes containing metakaolin (MK) and fly ash (FA). In this regard, three different algorithms have been used, namely multilayer perceptron (MLP), radial basis function (RBF), and support vector machine (SVR), to predict CS of concretes by considering most influencers input variables. These algorithms integrated with the grey wolf optimization (GWO) algorithm to increase the model's accuracy in predicting (GWMLP, GWRBF, and GWSVR). The proposed MLP models were implemented and evaluated in three different layers, wherein each layer, GWO, fitted the best neuron number of the hidden layer. Correspondingly, the key parameters of the SVR model are identified using the GWO method. Also, the optimization algorithm determines the hidden neurons' number and the spread value to set the RBF structure. The results show that the developed models all provide accurate predictions of the CS of concrete incorporating MK and FA with R² larger than 0.9972 and 0.9976 in the learning and testing stage, respectively. Regarding GWMLP models, the GWMLP1 model outperforms other GWMLP networks. All in all, GWSVR has the worst performance with the lowest indices, while the highest score belongs to GWRBF.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.29 no.10
• /
• pp.497-503
• /
• 2017

In recent years, studies on the prediction of building load using Artificial Neural Network (ANN) models have been actively conducted in the field of building energy In general, building loads predicted by ANN models show a sharp deviation unless large data sets are used for learning. On the other hands, some of the input data are hard to be acquired by common measuring devices. In this work, we estimate daily building loads with a limited number of input data and fewer pastdatasets (3 to 10 days). The proposed model with fewer input data gave satisfactory results as regards to the ASHRAE Guide Line showing 21% in CVRMSE and -3.23% in MBE. However, the level of accuracy cannot be enhanced since data used for learning are insufficient and the typical ANN models cannot account for thermal capacity effects of the building. An attempt proposed in this work is that learning procersses are sequenced frequrently and past data are accumulated for performance improvement. As a result, the model met the guidelines provided by ASHRAE, DOE, and IPMVP with by 17%, -1.4% in CVRMSE and MBE, respectively.

• Smart Structures and Systems
• /
• v.30 no.4
• /
• pp.421-432
• /
• 2022

Regarding the high significance of correct pan evaporation modeling, this study introduces two novel neuro-metaheuristic approaches to improve the accuracy of prediction for this parameter. Vortex search algorithms (VSA), sunflower optimization (SFO), and stochastic fractal search (SFS) are integrated with a multilayer perceptron neural network to create the VSA-MLPNN, SFO-MLPNN, and SFS-MLPNN hybrids. The climate data of Arcata-Eureka station (operated by the US environmental protection agency) belonging to the years 1986-1989 and the year 1990 are used for training and testing the models, respectively. Trying different configurations revealed that the best performance of the VSA, SFO, and SFS is obtained for the population size of 400, 300, and 100, respectively. The results were compared with a conventionally trained MLPNN to examine the effect of the metaheuristic algorithms. Overall, all four models presented a very reliable simulation. However, the SFS-MLPNN (mean absolute error, MAE = 0.0997 and Pearson correlation coefficient, R_P = 0.9957) was the most accurate model, followed by the VSA-MLPNN (MAE = 0.1058 and R_P = 0.9945), conventional MLPNN (MAE = 0.1062 and R_P = 0.9944), and SFO-MLPNN (MAE = 0.1305 and R_P = 0.9914). The findings indicated that employing the VSA and SFS results in improving the accuracy of the neural network in the prediction of pan evaporation. Hence, the suggested models are recommended for future practical applications.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.2
• /
• pp.1036-1043
• /
• 2014

The installation of PV system to the power distribution system is being increased as one of solutions for environmental pollution and energy crisis. Because the output efficiency of PV system is getting decreased because of the aging phenomenon and several operation obstacles, the technology development of output prediction and state diagnosis of PV modules are required in order to improve operation performance of PV modules. The conventional methods for output prediction by considering various parameters and standard test condition values of PV modules may have difficult and complex computation procedure and also their prediction values may produce large error. To overcome these problems, this paper proposes an optimal prediction algorithm and state diagnosis algorithm of PV modules by using least square methods of linear regression analysis. In addition, this paper presents a state diagnosis evaluation system of PV modules based on the proposed optimal algorithms of PV modules. From the simulation results of proposed evaluation system, it is confirmed that the proposed algorithms is a practical tool for state diagnosis of PV modules.

• International conference on construction engineering and project management
• /
• 2011.02a
• /
• pp.382-388
• /
• 2011

Being aware of the risk in advance necessitates intricate processes but is feasible. Although previous studies have demonstrated high accuracy, their performance still leaves room for improvement. A self-organizing feature map (SOM) based neurofuzzy model is developed in this study to provide another alternative for forecasting corporate financial distress. The model is designed to yield high prediction accuracy, as well as reference rules for evaluating corporate financial status. As a database, the study collects all financial reports from listed construction companies during the latest decade, resulting in over 1000 effective samples. The proportion of "failed" and "non-failed" companies is approximately 1:2. Each financial report is comprised of 25 ratios which are set as the input variable s. The proposed model integrates the concepts of pattern classification, fuzzy modeling and SOM-based optimization to predict corporate financial distress. The results exhibit a high accuracy rate at 85.1%. This model outperforms previous tools. A total of 97 rules are extracted from the proposed model which can be also used as reference for construction practitioners. Users may easily identify their corporate financial status by using these rules.