Il Chul Moon

Korea Advanced Institute of Science and Technology | KAIST · Department of Industrial and Systems Engineering

Training neural networks on a large dataset requires substantial computational costs. Dataset reduction selects or synthesizes data instances based on the large dataset, while minimizing the degradation in generalization performance from the full dataset. Existing methods utilize the neural network during the dataset reduction procedure, so the mod...

The collaborative filtering (CF) recommendation algorithm predicts the purchases of specific users based on their characteristics and purchase history. This study empirically analyzes the possibility of applying CF to the insurance industry using real customer data from South Korea. Using three different CF models, we examined the relevance of appl...

While the success of diffusion models has been witnessed in various domains, only a few works have investigated the variation of the generative process. In this paper, we introduce a new generative process that is closer to the reverse process than the original generative process, given the identical score checkpoint. Specifically, we adjust the ge...

Multi-resolution modeling (MRM) has been considered as an ideal form of simulation to acquire low-resolution scalability as well as high-resolution modeled details. Although both practical and theoretical interests exist in MRM, actual implementations were quite different in terms of cases and methods. Specifically, MRM implementations range from p...

Open-Set Domain Adaptation (OSDA) assumes that a target domain contains unknown classes, which are not discovered in a source domain. Existing domain adversarial learning methods are not suitable for OSDA because distribution matching with \textit{unknown} classes leads to the negative transfer. Previous OSDA methods have focused on matching the so...

Whereas diverse variations of diffusion models exist, expanding the linear diffusion into a nonlinear diffusion process is investigated only by a few works. The nonlinearity effect has been hardly understood, but intuitively, there would be more promising diffusion patterns to optimally train the generative distribution towards the data distributio...

Noisy labels are inevitable yet problematic in machine learning society. It ruins the generalization power of a classifier by making the classifier be trained to be overfitted to wrong labels. Existing methods on noisy label have focused on modifying classifier training procedure. It results in two possible problems. First, these methods are not ap...

Agent-based models (ABMs) highlight the importance of simulation validation, such as qualitative face validation and quantitative empirical validation. In particular, we focused on quantitative validation by adjusting simulation input parameters of the ABM. This study introduces an automatic calibration framework that combines the suggested dynamic...

Simulation has been applied to diverse domains such as urban growth modeling and market dynamics modeling. Some of these applications may require validations, based on some real-world observations modeled in the simulation. This validation can be conducted as either qualitative face-validation or quantitative empirical validation; however, as the i...

Searching for and comparing similar wafer maps can provide crucial information for root cause analysis in the manufacturing process of integrated circuits. Owing to the high dimensionality and complexity of defect patterns, comparison of similar maps in their entirety is inefficient. This paper proposes an automated similarity ranking system with a...

Recent advance in score-based models incorporates the stochastic differential equation (SDE), which brings the state-of-the art performance on image generation tasks. This paper improves such score-based models by analyzing the model at the zero perturbation noise. In real datasets, the score function diverges as the perturbation noise ($\sigma$) d...

The problem of fair classification can be mollified if we develop a method to remove the embedded sensitive information from the classification features. This line of separating the sensitive information is developed through the causal inference, and the causal inference enables the counterfactual generations to contrast the what-if case of the opp...

Attention computes the dependency between representations, and it encourages the model to focus on the important selective features. Attention-based models, such as Transformer and graph attention network (GAT), are widely utilized for sequential data and graph-structured data. This paper suggests a new interpretation and generalized structure of t...

Knowledge distillation is a method of transferring the knowledge from a pretrained complex teacher model to a student model, so a smaller network can replace a large teacher network at the deployment stage. To reduce the necessity of training a large teacher model, the recent literatures introduced a self-knowledge distillation, which trains a stud...

Tracking an object in a noisy environment is difficult, especially when unknown parameters affect the object’s behavior. In the case of a high-speed ballistic object, its trajectory is affected by changes in atmospheric conditions as well as various parameters of the object itself. To filter these latent factors of the dynamics model, this paper pr...

The recent development of likelihood-free inference aims training a flexible density estimator for the target posterior with a set of input-output pairs from simulation. Given the diversity of simulation structures, it is difficult to find a single unified inference method for each simulation model. This paper proposes a universally applicable regu...

The problem of fair classification can be mollified if we develop a method to remove the embedded sensitive information from the classification features. This line of separating the sensitive information is developed through the causal inference, and the causal inference enables the counterfactual generations to contrast the what-if case of the opp...

Active learning effectively collects data instances for training deep learning models when the labeled dataset is limited and the annotation cost is high. Besides active learning, data augmentation is also an effective technique to enlarge the limited amount of labeled instances. However, the potential gain from virtual instances generated by data...

Bayesian inference without the access of likelihood, called likelihood-free inference, is highlighted in simulation to yield a more realistic simulation result. Recent research updates an approximate posterior sequentially with the cumulative simulation input-output pairs over inference rounds. This paper observes that previous algorithms with Mont...

Recently, the population of Seoul has been affected by particulate matter in the atmosphere. This problem can be addressed by developing an elaborate forecasting model to estimate the concentration of fine dust in the metropolitan area. We present a forecasting model of the fine dust concentration with an extended range of input variables, compared...

Attention compute the dependency between representations, and it encourages the model to focus on the important selective features. Among the attention methods, the scaled dot-product attention is widely utilized in many models. This paper suggests a generalized structure of the scaled dot-product attention with similarity and magnitude terms. We d...

This paper proposes Dirichlet Variational Autoencoder (DirVAE) using a Dirichlet prior. To infer the parameters of DirVAE, we utilize the stochastic gradient method by approximating the inverse cumulative distribution function of the Gamma distribution, which is a component of the Dirichlet distribution. This approximation on a new prior led an inv...

Generative Adversarial Network (GAN) can be viewed as an implicit estimator of a data distribution, and this perspective motivates using GAN in the true parameter estimation under a complex black-box generative model. While previous works investigated how to backpropagate gradients through the black-box model, this paper suggests an augmented neura...

Recent researches demonstrate that word embeddings, trained on the human-generated corpus, have strong gender biases in embedding spaces, and these biases can result in the prejudiced results from the downstream tasks, i.e. sentiment analysis. Whereas the previous debiasing models project word embeddings into a linear subspace, we introduce a Laten...

The success of military operations depends on soldiers’ execution of the operation as well as resources used for the operation. However, this does not mean that more men and firepower will ensure victory. Military units, just like any other organization, are collections of distributed elements, and improving the organization or command and control...

Recent studies identified that sequential Recommendation is improved by the attention mechanism. By following this development, we propose Relation-Aware Kernelized Self-Attention (RKSA) adopting a self-attention mechanism of the Transformer with augmentation of a probabilistic model. The original self-attention of Transformer is a deterministic me...

While simulations have been utilized in diverse domains, such as urban growth modeling, market dynamics modeling, etc; some of these applications may require validations based upon some real-world observations modeled in the simulation, as well. This validation has been categorized into either qualitative face-validation or quantitative empirical v...

Successful application processing sequential data, such as text and speech, requires an improved generalization performance of recurrent neural networks (RNNs). Dropout techniques for RNNs were introduced to respond to these demands, but we conjecture that the dropout on RNNs could have been improved by adopting the adversarial concept. This paper...

A long user history inevitably reflects the transitions of personal interests over time. The analyses on the user history require the robust sequential model to anticipate the transitions and the decays of user interests. The user history is often modeled by various RNN structures, but the RNN structures in the recommendation system still suffer fr...

Long Short-Term Memory (LSTM) infers the long term dependency through a cell state maintained by the input and the forget gate structures, which models a gate output as a value in [0,1] through a sigmoid function. However, due to the graduality of the sigmoid function, the sigmoid gate is not flexible in representing multi-modality or skewness. Bes...

Understanding politics is challenging because the politics take the influence from everything. Even we limit ourselves to the political context in the legislative processes; we need a better understanding of latent factors, such as legislators, bills, their ideal points, and their relations. From the modeling perspective, this is difficult 1) becau...

A long user history inevitably reflects the transitions of personal interests over time. The analyses on the user history require the robust sequential model to anticipate the transitions and the decays of user interests. The user history is often modeled by various RNN structures, but the RNN structures in the recommendation system still suffer fr...

Successful application processing sequential data, such as text and speech, requires an improved generalization performance of recurrent neural networks (RNNs). Dropout techniques for RNNs were introduced to respond to these demands, but we conjecture that the dropout on RNNs could have been improved by adopting the adversarial concept. This paper...

The joint optimization of representation learning and clustering in the embedding space has experienced a breakthrough in recent years. In spite of the advance, clustering with representation learning has been limited to flat-level categories, which often involves cohesive clustering with a focus on instance relations. To overcome the limitations o...

This paper proposes Dirichlet Variational Autoencoder (DirVAE) using a Dirichlet prior for a continuous latent variable that exhibits the characteristic of the categorical probabilities. To infer the parameters of DirVAE, we utilize the stochastic gradient method by approximating the Gamma distribution, which is a component of the Dirichlet distrib...

This paper proposes a deep reinforcement learning approach in order to optimize a sequence of tasks efficiently with the aid of image processing techniques used in computer vision. The proposed algorithm can be employed to solve the traveling salesman problem (TSP), a combinatorial optimization problem that determines the optimum trajectory of city...

This paper proposes a new method of predicting the future state of a ballistic target trajectory. There have been a number of estimation methods that utilize the variations of Kalman filters, and the prediction of the future states followed the simple propagations of the target dynamic equations. However, these simple propagations suffered from no...

Defense modeling and simulation (DM&S) has brought insights into how to efficiently operate combat entities, such as soldiers and weapon systems. Most DM&S works have been developed to reflect accurate descriptions of military doctrines, yet these doctrines provide only guidelines of military operations, not details about how the combat entities sh...

Identification of prescription patterns is a useful and interesting goal from multiple perspectives. The identified prescription patterns may expand the horizons of medical knowledge, and may be evaluated by subject matter experts to label certain patterns as anomalies calling for further investigation; for example, in prescription costs for insura...

Recommender systems offer critical services in the age of mass information. A good recommender system selects a certain item for a specific user by recognizing why the user might like the item. This awareness implies that the system should model the background of the items and the users. This background modeling for recommendation is tackled throug...

Wide variance exists among individuals and institutions for treating patients with medicine. This paper analyzes prescription patterns using a topic model with more than four million prescriptions. Specifically, we propose the disease-medicine pattern model (DMPM) to extract patterns from a large collection of insurance data by considering disease...

Recently, the training with adversarial examples, which are generated by adding a small but worst-case perturbation on input examples, has been proved to improve generalization performance of neural networks. In contrast to the individually biased inputs to enhance the generality, this paper introduces adversarial dropout, which is a minimal set of...

Many disasters have occurred around the world and have caused sizable damage. A disaster, called a mass casualty incident (MCI), generates a large number of casualties that overwhelm the capacity of local medical resources, and the disaster responses to the MCI requires many interactions among the disaster responders. To evaluate the efficiency of...

Despite the proliferation of topic models, the organization of topics from the probabilistic models needs improvement in two ways: the better structured presentation of topics and the incorporation of domain knowledge on the corpus. The structured presentation, i.e., the hierarchical topic model, helps in categorizing similar topics; the incorporat...

This thesis introduces a case study of utilizing the decentralized partially observableMarkov decision process (DEC-POMDP) in modeling information delivery behavior under the network centric warfare settings. The deployed troops are modeled as bounded rational agents, and they have different communication success possibilities in the long range and...

Modeling combat behavior is an important, yet complicated task because the combat behavior emerges from the rationality as well as the irrationality. For instance, when a soldier confronts a dilemma on accomplishing his mission and saving his life, it is difficult to model his ongoing thoughts with a simple model. This paper presents (1) how to rec...

Discrete event models are widely used to replicate, analyze, and understand complex systems. DEVS (Discrete Event System Specification) formalism enables hierarchical modeling, so it provides an efficiency in the model development of complex models. However, the hierarchical modeling incurs prolonged simulation executions due to indirect event exch...

The purpose of this study was to investigate consumers` perceptions of sodium saccharin in social media. Data was collected from Naver blogs and Naver web communities (Korean representative portal web-site), and media reports including comment sections on a Yonhap news website (Korean largest news agency). The results from Naver blogs and Naver web...

Modelling command and control (C2) is regarded as a difficult task because of the complexity of the decision-making requiredby individuals in combat. Despite the difficulties, C2 modelling is frequently used for high echelon units, i.e. battalion, divisionand above. This paper extends these models to the lowest army unit: the infantry company. Prev...

Abstract A series of events generates multiple types of time series data, such as numeric and text data over time, and the variations of the data types capture the events from different angles. This paper aims to integrate the analyses on such numerical and text time-series data influenced by common events with a single model to better understand t...

Disaster response operations are critical for decreasing the devastating impacts that result in casualties and property damages. Since these operations require cooperation in dynamic and complex situations, the responding organizations require a solid organizational structure collectively. This article introduces computational designs and evaluatio...

From a military standpoint, a river is an area that should be avoided in a potential engagement because of lack of cover and the necessity of dividing the unit while crossing. Thus, a key point of a river-crossing operation is speed. Many efforts have been made to enable faster river crossing by improvement of tactics, techniques, and procedures (T...

To reduce overpopulation around Seoul, Korea, the government implemented a relocation policy of public officers by moving the government complex. This implies that there will be a negative impact on the suburban area that originally hosted the complex, but we do not know the magnitude of the impact. Therefore, this paper presents a micro-level esti...

Recent research indicates that a sentiment lexicon focusing on a specific domain leads to better sentiment analyses compared to a general-purpose sentiment lexicon, such as Senti-WordNet. In spite of this potential improvement, the cost of building a domain-specific sentiment lexicon hinders its wider and more practical applications. To compensate...

Agent-based models are generative models to provide the unrecorded, yet important storylines to model users. Such generative models might have less accuracy in the prediction, but the provided insights would be invaluable even compared to the accurate predictions. In spite of this different value proposition, outsiders as well as insiders of the ag...

As agent-based models (ABMs) are applied to various domains, the efficiency of model development has become an important issue in its applications. The current practice is that many models are developed from scratch, while they could have been built by reusing existing models. Moreover, when models need reconfiguration, they often need to be rebuil...

Recently, a challenge in defence modelling and simulation is that simulating a satisfactory number of scenarios often requires an infeasible runtime. This paper resolves this challenge by utilizing a tabulation technique that encourages reuses of the previous simulation results in hierarchical models. For example, a mission-level model may contain...

The bombardment of a metropolis is considered a nightmare scenario. To reduce losses from such an assault, big cities have developed evacuation policies in case of bombardment. However, to build efficient evacuation policies, much footing data is required that considers both military and civilian views. Agent-based modeling and simulation could be...

Analyzing patient records is important for improving the quality of medical services and for understanding each patient's historical diseases. However, the huge size of the data requires statistical analysis procedures. In this paper, we proposed a probabilistic model-the disease-medicine topic model (DMTM)-to explore connected knowledge about dise...

Understanding hospitals' relationships is critical to the analysis of public healthcare environment. There have been many attempts to analyze medical environment at a personal level. Recently, at an organizational level, there has been some advance in research into examining a relationship between hospitals. However, the formation of linkages is re...

Disasters and responses have evolved over-time, and the evolution has been affected by various factors, such as societal change, climate change, and technological advance. To better prepare the future disasters, we need to estimate the evolution trend of the past disasters and the responses. This paper analyzes the academic articles of the field wi...

Under urban crisis situations, one of the most important response tasks is routing the vehicles. Such crisis initiates a massive evacuation from the disaster scene to the outside; at the same time, crisis responders have to enter the scene. Whereas we need routes for responders, current disaster response plans frequently dictate to turn bidirection...

Understanding the emergence of collective emotions is critical to the analysis of online and offline societies. The agent-based simulation community has developed various social norm models to see the polarization of collective emotions. Yet, a few models have psychological background as fundamentals, as well as statistical validation, and this pap...

Power TAC (Power Trading Agent Competition) is an agent-based simulation for competitions between electricity brokering agents on the smart grid. To win the competition, agents obtain electricity from the electricity wholesale market among the power plants. In this operation, a key to success is balancing the demand of the customer and the supply f...

Combat Modeling and Simulation (M&S) is significant to decision makers who predict the next direction of wars. Classical methodologies for combat M&S aimed to describe the exact behaviors of combat entities from military doctrines, yet they had a limitation of describing reasonable behaviors of combat entities that did not appear in the doctrines....

Modeling and simulating a real world scenario is fundamentally an abstraction that takes only part of the given scenario into the model. Furthermore, the level of detail in the model, a.k.a. the resolution, plays an important role in the modeling and simulation process. Finally, the abstraction and resolution of the model determine the fidelity of...

To reduce overpopulation around Seoul, Korea, the government implemented a relocation policy of public officers by moving the government complex. This implies that there will be a negative impact to the suburban area that originally hosted the complex, but we do not know the magnitude of the impact. Therefore, this paper presents a micro-level esti...

As the complexity of military operations increases, the defense modeling and simulation (DM&S) has contributed in analytically improving doctrines at the engineering, engagement, mission, and campaign levels. To date, defense modelers concentrate on the best representation of their targeted system at their targeted modeling level, and the modelers...