• Journal of Broadcast Engineering
• /
• v.28 no.3
• /
• pp.275-284
• /
• 2023

Gene expression data can be utilized in various studies, including the prediction of disease prognosis. However, there are challenges associated with collecting enough data due to cost constraints. In this paper, we propose a gene expression data generation model based on Conditional Variational Autoencoder. Our results demonstrate that the proposed model generates synthetic data with superior quality compared to two other state-of-the-art models for gene expression data generation, namely the Wasserstein Generative Adversarial Network with Gradient Penalty based model and the structured data generation models CTGAN and TVAE.

• Smart Media Journal
• /
• v.10 no.1
• /
• pp.79-87
• /
• 2021

In recent years research on automatic font generation with machine learning mainly focus on using transformation-based methods, in comparison, generative model-based methods of font generation have received less attention. Transformation-based methods learn a mapping of the transformations from an existing input to a target. This makes them ambiguous because in some cases a single input reference may correspond to multiple possible outputs. In this work, we focus on font generation using the generative model-based methods which learn the buildup of the characters from noise-to-image. We propose a novel way to train a conditional generative deep neural model so that we can achieve font style control on the generated font images. Our research demonstrates how to generate new font images conditioned on both character class labels and character style labels when using the generative model-based methods. We achieve this by introducing a modified generator network which is given inputs noise, character class, and style, which help us to calculate losses separately for the character class labels and character style labels. We show that adding the character style vector on top of the character class vector separately gives the model rich information about the font and enables us to explicitly specify not only the character class but also the character style that we want the model to generate.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.37 no.3
• /
• pp.135-150
• /
• 2012

This paper presents a portfolio model for a long-term power generation mix problem. The proposed portfolio model evaluates generation mix by considering the tradeoffs between the expected cost for power generation and its variability. Unlike conventional portfolio models measuring variance, we introduce Conditional Value-at-Risk (CVaR) in designing the variability with aims to considering events that are enormously expensive but are rare such as nuclear power plant accidents. Further, we consider uncertainties associated with future electricity demand, fuel prices and their correlations, and capital costs for power plant investments. To obtain an objective generation by each energy source, we employ the sample average approximation method that approximates the stochastic objective function by taking the average of large sample values so that provides asymptotic convergence of optimal solutions. In addition, the method includes Monte Carlo simulation techniques in generating random samples from multivariate distributions. Applications of the proposed model and method are demonstrated through a case study of an electricity industry with nuclear, coal, oil (OCGT), and LNG (CCGT) in South Korea.

• The Journal of Korea Robotics Society
• /
• v.4 no.1
• /
• pp.56-61
• /
• 2009

A probability-based data generation is a typical context-generation method that is a not only simple and strong data generation method but also easy to update generation conditions. However, the probability-based context-generation method has been found its natural-born ambiguousness and confliction problems in generated context data. In order to compensate for the disadvantages of the probabilistic random data generation method, a situation propagation network is proposed in this paper. The situation propagating network is designed to update parameters of probability functions are included in probability-based data generation model. The proposed probability-based context-generation model generates two kinds of contexts: one is related to independent contexts, and the other is related to conditional contexts. The results of the proposed model are compared with the results of the probabilitybased model with respect to performance, reduction of ambiguity, and confliction.

• Annual Conference of KIPS
• /
• 2021.11a
• /
• pp.764-767
• /
• 2021

While Generative Adversarial Networks (GANs) have seen huge success in image synthesis tasks, synthesizing high-quality images from text descriptions is a challenging problem in computer vision. This paper proposes a method named Text-to-Face Generation Using Multi-Scale Gradients for Conditional Generative Adversarial Networks (T2F-MSGGANs) that combines GANs and a natural language processing model to create human faces has features found in the input text. The proposed method addresses two problems of GANs: model collapse and training instability by investigating how gradients at multiple scales can be used to generate high-resolution images. We show that T2F-MSGGANs converge stably and generate good-quality images.

• The Korean Journal of Applied Statistics
• /
• v.10 no.1
• /
• pp.151-161
• /
• 1997

An easy Monte Carlo Gibbs sampling approach is suggested for Bayesian analysis of cumulative logit models for ordinal polytomous data. Because in the cumulative logit model the posterior conditional distributions of parameters are not given in convenient forms for random sample generation, appropriate latent variables are introduced into the model so that in the new model all the conditional distributions are given in very convenient forms for implementation of the Gibbs sampler.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.5
• /
• pp.217-223
• /
• 2023

In this paper, we propose a new model for the conditional generation of music, considering key and rhythm, fundamental elements of music. MIDI sheet music is converted into a WAV format, which is then transformed into a Mel Spectrogram using the Short-Time Fourier Transform (STFT). Using this information, key and rhythm details are classified by passing through two Convolutional Neural Networks (CNNs), and this information is again fed into the Music Transformer. The key and rhythm details are combined by differentially multiplying the weights and the embedding vectors of the MIDI events. Several experiments are conducted, including a process for determining the optimal weights. This research represents a new effort to integrate essential elements into music generation and explains the detailed structure and operating principles of the model, verifying its effects and potentials through experiments. In this study, the accuracy for rhythm classification reached 94.7%, the accuracy for key classification reached 92.1%, and the Negative Likelihood based on the weights of the embedding vector resulted in 3.01.

• KIPS Transactions on Software and Data Engineering
• /
• v.3 no.10
• /
• pp.407-420
• /
• 2014

In general, the Korean morpheme analysis procedure is divided into two steps. In the first step as an ambiguity generation step, an Eojeol is analyzed into many morpheme sequences as candidates. In the second step, one appropriate candidate is chosen by using contextual information. Hidden Markov Model(HMM) is typically applied in the second step. This paper proposes Sub-word Conditional Probability(SCP) model as an alternate algorithm. SCP uses sub-word information of adjacent eojeol first. If it failed, then SCP use morpheme information restrictively. In the accuracy and speed comparative test, HMM's accuracy is 96.49% and SCP's accuracy is just 0.07% lower. But SCP reduced processing time 53%.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.2 no.3
• /
• pp.29-39
• /
• 1992

We investigate a systematic design procedure of automatic rule generation of fuzzy logic based controllers for highly nonlinear dynamic systems such as an engine dynamic modle. By "automatic rule generation" we mean autonomous clustering or collection of such meaningful transitional relations from one conditional subspace to another. During the design procedure, we also consider optimaly control strategies such as minimum squared error, near minimum time, minimum energy or combined performance critiera. Fuzzy feedback control systems designed by our method have the properties of closed-loop stability, robustness under parameter variabitions, and a certain degree of optimality. Most of all, the main advantage of the proposed approach is that reliability can be potentially increased even if a large grain of uncertainty is involved within the control system under consideration. A numerical example is shown in which we apply our strategic fuzzy controller dwsign to a highly nonlinear model of engine idling speed control.d control.

• The Transactions of the Korea Information Processing Society
• /
• v.2 no.2
• /
• pp.254-263
• /
• 1995

In this paper we discuss Moore and Mealy-model Time-stationary control schemes of pipelined data paths of Application Specific, Integrated Circuits (ASICs). We developed a method to synthesize both a Moore and a Mealy-style Finite State Machine(FSM) controller specifications given a pipelined data path with conditional branches. The control synthesis task consists of the generation of control specification and the FSM synthesis. The control specification procedure generates a FSM specification in the form of a state table. The different partitioning schemes are applied to each FSM controller so as to minimize the total area. Experimental results show the characteristics of the two different control styles and the effects of these two models on cost and performance.