• The Journal of the Acoustical Society of Korea
• /
• v.28 no.6
• /
• pp.572-577
• /
• 2009

Large corpus-based concatenating Text-to-Speech (TTS) systems can generate natural synthetic speech without additional signal processing. To prune the redundant speech segments in a large speech segment DB, we can utilize a decision-tree based triphone clustering algorithm widely used in speech recognition area. But, the conventional methods have problems in representing the acoustic transitional characteristics of the phones and in applying context questions with hierarchic priority. In this paper, we propose a new clustering algorithm to downsize the speech DB. Firstly, three 13th order MFCC vectors from first, medial, and final frame of a phone are combined into a 39 dimensional vector to represent the transitional characteristics of a phone. And then the hierarchically grouped three question sets are used to construct the triphone trees. For the performance test, we used DTW algorithm to calculate the acoustic similarity between the target triphone and the triphone from the tree search result. Experimental results show that the proposed method can reduce the size of speech DB by 23% and select better phones with higher acoustic similarity. Therefore the proposed method can be applied to make a small sized TTS.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.19 no.2
• /
• pp.1-11
• /
• 2024

In the midst of the current turbulent global economy, traditional investment metrics are undergoing a metamorphosis, signaling the onset of what's often referred to as an "Investment cold season". Early-stage startups, despite their boundless potential, grapple with immediate revenue constraints, intensifying their pursuit of critical investments. While financial indicators once took center stage in investment evaluations, a notable paradigm shift is underway. Organizational culture, once relegated to the sidelines, has now emerged as a linchpin in forecasting a startup's resilience and enduring trajectory. Our comprehensive research, integrating insights from CVF and OCAI, unveils the intricate relationship between organizational culture and its magnetic appeal to investors. The results indicate that startups with a pronounced external focus, expertly balanced with flexibility and stability, hold particular allure for investment consideration. Furthermore, the study underscores the pivotal role of adhocracy and market-driven mindsets in shaping investment desirability. A significant observation emerges from the study: startups, whether they secured investment or failed to do so, consistently display strong clan culture, highlighting the widespread importance of nurturing a positive employee environment. Leadership deeply anchored in market culture, combined with an unwavering commitment to innovation and harmonious organizational practices, emerges as a potent recipe for attracting investor attention. Our model, with an impressive 88.3% predictive accuracy, serves as a guiding light for startups and astute investors, illuminating the intricate interplay of culture and investment success in today's economic landscape.

• Applied Chemistry for Engineering
• /
• v.35 no.1
• /
• pp.54-60
• /
• 2024

The present study was aimed to evaluate the removal of the trace pollutants (heavy metals and microplastics) in the sewage treatment plant by using the jellyfish Extract at Immunity reaction (JEI) of Aurelia coerulea. The experiment was conducted on two different scales: the lab scale using a Jar-tester and the Pilot system scale equipped with two newly developed devices in the laboratory, the active tube connection mixed system and the concentration integrated separation device. Compared to anionic polymers currently used in the field, JEI showed similar or higher efficiency to remove the trace pollutants. When JEI was added to the effluent through the Pilot system, the combination of JEI and the trace pollutants was maximized through two mixing processes, and as a result, the removal rate of the trace pollutants was greatly improved. Based on these results, we propose the present technology as an alternative to removing trace pollutants that can reduce ecosystem risk and minimize the generation of inorganic waste, away from the existing method.

• Journal of Korea Water Resources Association
• /
• v.57 no.2
• /
• pp.73-85
• /
• 2024

Predicting water quality of rivers and reservoirs is necessary for the management of water resources. Artificial Neural Networks (ANNs) have been used in many studies to predict water quality with high accuracy. Previous studies have used Gradient Descent (GD)-based optimizers as an optimizer, an operator of ANN that searches parameters. However, GD-based optimizers have the disadvantages of the possibility of local optimal convergence and absence of a solution storage and comparison structure. This study developed improved optimizers to overcome the disadvantages of GD-based optimizers. Proposed optimizers are optimizers that combine adaptive moments (Adam) and Nesterov-accelerated adaptive moments (Nadam), which have low learning errors among GD-based optimizers, with Harmony Search (HS) or Novel Self-adaptive Harmony Search (NSHS). To evaluate the performance of Long Short-Term Memory (LSTM) using improved optimizers, the water quality data from the Dasan water quality monitoring station were used for training and prediction. Comparing the learning results, Mean Squared Error (MSE) of LSTM using Nadam combined with NSHS (NadamNSHS) was the lowest at 0.002921. In addition, the prediction rankings according to MSE and R² for the four water quality indices for each optimizer were compared. Comparing the average of ranking for each optimizer, it was confirmed that LSTM using NadamNSHS was the highest at 2.25.

• Journal of Applied Biological Chemistry
• /
• v.66
• /
• pp.402-407
• /
• 2023

The regulation of platelet aggregation is crucial for maintaining normal hemostasis, but abnormal or excessive platelet aggregation can contribute to cardiovascular disorders such as stroke, atherosclerosis, and thrombosis. Therefore, identifying substances that can control or suppress platelet aggregation is a promising approach for the prevention and treatment of these conditions. Artemisinin, a compound derived from Artemisia or Scopolia plants, has shown potential in various areas such as anticancer and Alzheimer's disease research. However, the specific role and mechanisms by which artemisinin influences platelet activation and thrombus formation are not yet fully understood. This study investigated the effects of artemisinin on platelet activation and thrombus formation. As a result, cAMP production were increased significantly by artemisinin, as well as phosphorylated VASP and IP₃R which are substrates to cAMP-dependent kinase by artemisinin in a significant manner. The Ca²⁺ normally mobilized from the dense tubular system was inhibited due to IP₃R phosphorylation from artemisinin, and phosphorylated VASP by artemisinin aided in inhibiting platelet activity via αIIb/β3 platelet membrane inactivation and inhibiting fibrinogen binding. Finally, artemisinin inhibited thrombin-induced thrombus formation. Therefore, we suggest that artemisinin has importance with cardiovascular diseases stemming from the abnormal platelets activation and thrombus formation by acting as an effective prophylactic and therapeutic agent.

• Applied Chemistry for Engineering
• /
• v.35 no.3
• /
• pp.192-199
• /
• 2024

In this study, carbamazepine (CBZ) imprinted starch/PVA-based biomaterials were prepared by the casting method and UV irradiation, and their physicochemical properties, CBZ adsorption ability, and release properties were investigated. The surface properties of the prepared biomaterials were characterized using FE-SEM, while the stability of CBZ under UV irradiation and the functional groups of the biomaterials were characterized using FT-IR analysis. The adsorption properties of CBZ on the biomaterials were evaluated by binding isotherm and Scatchard plot. Results indicate that CBZ imprinted biomaterials possess a specific binding site of CBZ. To evaluate the applicability of the transdermal drug delivery system, the release properties of CBZ from prepared biomaterials using various pH buffers and artificial skin at 36.5 ℃ were investigated. Results indicated that the CBZ release at high pH was faster than at low pH. In addition, CBZ was released continuously for 12 h in the artificial skin test. The drug release mechanism of CBZ followed a pseudo-Fickian diffusion mechanism in buffer solution, whereas the release from artificial skin exhibited a non-Fickian diffusion mechanism.

• KOREAN JOURNAL OF PACKAGING SCIENCE & TECHNOLOGY
• /
• v.30 no.2
• /
• pp.101-106
• /
• 2024

Current cold chain logistics relying on organic or eutectic materials within the 2~8℃ range as secondary fluids often face limitations in heat storage capacity, necessitating high energy consumption and large volume capacity. An effective approach to address this challenge is by incorporating polymers to enhance the heat storage capacity of eutectic materials. In this study, we investigated the impact of polyethylene glycols (PEGs) on phase change materials using Fourier transform infrared spectroscopy (FT-IR), differential scanning calorimeter (DSC), analyses of endothermic and exothermic phase change processes, and an accelerated thermal cycling test. Our findings indicate that the introduction of PEGs into the phase change materials can lead to improvements in latent heat, thermal conductivity, and 2~8℃ retention time. This enhancement is attributed to the high latent heat and thermal conductivity of the polymer, along with its ability to inhibit crystal formation in the eutectic mixture.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.9
• /
• pp.1-7
• /
• 2024

In this paper, we proposes a Convolutional Neural Networks(CNN) equipped with Batch Normalization(BN) for handwritten digit recognition training the MNIST dataset. Aiming to surpass the performance of LeNet-5 by LeCun et al., a 6-layer neural network was designed. The proposed model processes 28×28 pixel images through convolution, Max Pooling, and Fully connected layers, with the batch normalization to improve learning stability and performance. The experiment utilized 60,000 training images and 10,000 test images, applying the Momentum optimization algorithm. The model configuration used 30 filters with a 5×5 filter size, padding 0, stride 1, and ReLU as activation function. The training process was set with a mini-batch size of 100, 20 epochs in total, and a learning rate of 0.1. As a result, the proposed model achieved a test accuracy of 99.22%, surpassing LeNet-5's 99.05%, and recorded an F1-score of 0.9919, demonstrating the model's performance. Moreover, the 6-layer model proposed in this paper emphasizes model efficiency with a simpler structure compared to LeCun et al.'s LeNet-5 (7-layer model) and the model proposed by Ji, Chun and Kim (10-layer model). The results of this study show potential for application in real industrial applications such as AI vision inspection systems. It is expected to be effectively applied in smart factories, particularly in determining the defective status of parts.

• Applied Biological Chemistry
• /
• v.9
• /
• pp.41-57
• /
• 1968

Changes of acids (total, titratable and combined form) and sugars (total, reducing and non-reducing) in the edible part and the rind of 17 varieties the in growing and ripening period were investigated. The results were summarized as follows. 1) The percentage of rind was notably decreased in growing period and slightly in the ripening period- It may suggest that the rates of translocation of metabolite from leaves to each part of fruit are different with growth phase. 2) The heavier the weight of fruit, the higher the percentage of rind was and the varieties having over 200 g per fruit showed the value over 30 in the rind percentage and over 15 in the number seeds per fruit. 3) Total acid contents in the rind were highest at the maximum grow th of fruit except in Citrus grandis having tie lowest value (below 20 me/100 g F.W). of total acid at maximum point in which total acid content is steadily increase. 4) Total acid and titratable acid in the edible part and total acid and combined acid in whole fruit life showed 0.933 and 0.970 of correlation coefficient significant at 1% level respectively, and most acid in the edible part was titratable acid(73%) whereas acid in the rind consists mostly of combined acid. 5) The content of combined acid in the ripening period increased in the edible part and decreased in the rind. It may be contributed to translocation of some cations from the rind to the edible part. 6) The grouping criteria on citrus fruit were applicable on melon, watermelon and tomatoes. 7) The contents of total sugar and non-reducing sugar in the edible part were continuously increased whereas the content of reducing sugar were decreased in certain varieties, notablly in citrus natsudaidai. The correlation coefficient between total sugar and reducing sugar in the edible part with ripening decreased as $0.849^{**},\;0.732^{**}.\;0.583^*$. ( $^{**}$: significant at 1% level and $^{*}:$: at 5%) 8) 61% of total sugar in the edible part was non-reducing sugar whereas 88% of total sugar in the rind was reducing form at the end of ripening and the correlation coefficient between total and non-reducing sugar in the edible part was 0.861 end total and reducing sugar in the rind was 0.972, both significant at 1% level. 9) Varieties having the percentage of the rind below 36 showed higher value than I in the ratio of total sugar in the edible part to one in the rind. It may suggest that there exists any intimate relation between relative sugar content and growth rate of fruit parts. 10) Citrus unshiu in Guje island showed lower values in the content of acid and sugar, and the rind percentage but higher sweetness index (the ratio of total sugar to titratable acid) comparing with the same variety in Jeiu.

• Journal of Intelligence and Information Systems
• /
• v.25 no.2
• /
• pp.39-55
• /
• 2019

Recently banks and large financial institutions have introduced lots of Robo-Advisor products. Robo-Advisor is a Robot to produce the optimal asset allocation portfolio for investors by using the financial engineering algorithms without any human intervention. Since the first introduction in Wall Street in 2008, the market size has grown to 60 billion dollars and is expected to expand to 2,000 billion dollars by 2020. Since Robo-Advisor algorithms suggest asset allocation output to investors, mathematical or statistical asset allocation strategies are applied. Mean variance optimization model developed by Markowitz is the typical asset allocation model. The model is a simple but quite intuitive portfolio strategy. For example, assets are allocated in order to minimize the risk on the portfolio while maximizing the expected return on the portfolio using optimization techniques. Despite its theoretical background, both academics and practitioners find that the standard mean variance optimization portfolio is very sensitive to the expected returns calculated by past price data. Corner solutions are often found to be allocated only to a few assets. The Black-Litterman Optimization model overcomes these problems by choosing a neutral Capital Asset Pricing Model equilibrium point. Implied equilibrium returns of each asset are derived from equilibrium market portfolio through reverse optimization. The Black-Litterman model uses a Bayesian approach to combine the subjective views on the price forecast of one or more assets with implied equilibrium returns, resulting a new estimates of risk and expected returns. These new estimates can produce optimal portfolio by the well-known Markowitz mean-variance optimization algorithm. If the investor does not have any views on his asset classes, the Black-Litterman optimization model produce the same portfolio as the market portfolio. What if the subjective views are incorrect? A survey on reports of stocks performance recommended by securities analysts show very poor results. Therefore the incorrect views combined with implied equilibrium returns may produce very poor portfolio output to the Black-Litterman model users. This paper suggests an objective investor views model based on Support Vector Machines(SVM), which have showed good performance results in stock price forecasting. SVM is a discriminative classifier defined by a separating hyper plane. The linear, radial basis and polynomial kernel functions are used to learn the hyper planes. Input variables for the SVM are returns, standard deviations, Stochastics %K and price parity degree for each asset class. SVM output returns expected stock price movements and their probabilities, which are used as input variables in the intelligent views model. The stock price movements are categorized by three phases; down, neutral and up. The expected stock returns make P matrix and their probability results are used in Q matrix. Implied equilibrium returns vector is combined with the intelligent views matrix, resulting the Black-Litterman optimal portfolio. For comparisons, Markowitz mean-variance optimization model and risk parity model are used. The value weighted market portfolio and equal weighted market portfolio are used as benchmark indexes. We collect the 8 KOSPI 200 sector indexes from January 2008 to December 2018 including 132 monthly index values. Training period is from 2008 to 2015 and testing period is from 2016 to 2018. Our suggested intelligent view model combined with implied equilibrium returns produced the optimal Black-Litterman portfolio. The out of sample period portfolio showed better performance compared with the well-known Markowitz mean-variance optimization portfolio, risk parity portfolio and market portfolio. The total return from 3 year-period Black-Litterman portfolio records 6.4%, which is the highest value. The maximum draw down is -20.8%, which is also the lowest value. Sharpe Ratio shows the highest value, 0.17. It measures the return to risk ratio. Overall, our suggested view model shows the possibility of replacing subjective analysts's views with objective view model for practitioners to apply the Robo-Advisor asset allocation algorithms in the real trading fields.