Eunji Park

Eun-Ji Park, Sangwon Yun

Recent studies have proposed interpreting the training process from an ergodic perspective. Building on this foundation, we present a unified framework for understanding and accelerating the training of deep neural networks via stochastic gradient descent (SGD). By analyzing the geometric landscape of the objective function we introduce a practical diagnostic, the running estimate of the largest Lyapunov exponent, which provably distinguishes genuine convergence toward stable minimizers from mere statistical stabilization near saddle points. We then propose a ghost category extension for standard classifiers that adds auxiliary ghost output nodes so the model gains extra descent directions that open a lateral corridor around narrow loss barriers and enable the optimizer to bypass poor basins during the early training phase. We show that this extension strictly reduces the approximation error and that after sufficient convergence the ghost dimensions collapse so that the extended model coincides with the original one and there exists a path in the enlarged parameter space along which the total loss does not increase. Taken together, these results provide a principled architecture level intervention that accelerates early stage trainability while preserving asymptotic behavior and simultaneously serves as an architecture-friendly regularizer.

Young Ah Choi, Kyung Ho Park, Eunji Park et al.

In the era of intelligent transportation, driver behavior profiling has become a beneficial technology as it provides knowledge regarding the driver's aggressiveness. Previous approaches achieved promising driver behavior profiling performance through establishing statistical heuristics rules or supervised learning-based models. Still, there exist limits that the practitioner should prepare a labeled dataset, and prior approaches could not classify aggressive behaviors which are not known a priori. In pursuit of improving the aforementioned drawbacks, we propose a novel approach to driver behavior profiling leveraging an unsupervised learning paradigm. First, we cast the driver behavior profiling problem as anomaly detection. Second, we established recurrent neural networks that predict the next feature vector given a sequence of feature vectors. We trained the model with normal driver data only. As a result, our model yields high regression error given a sequence of aggressive driver behavior and low error given at a sequence of normal driver behavior. We figured this difference of error between normal and aggressive driver behavior can be an adequate flag for driver behavior profiling and accomplished a precise performance in experiments. Lastly, we further analyzed the optimal level of sequence length for identifying each aggressive driver behavior. We expect the proposed approach to be a useful baseline for unsupervised driver behavior profiling and contribute to the efficient, intelligent transportation ecosystem.

Eunji Park, Kyung Ho Park, Huy Kang Kim

The game industry has long been troubled by malicious activities utilizing game bots. The game bots disturb other game players and destroy the environmental system of the games. For these reasons, the game industry put their best efforts to detect the game bots among players' characters using the learning-based detections. However, one problem with the detection methodologies is that they do not provide rational explanations about their decisions. To resolve this problem, in this work, we investigate the explainabilities of the game bot detection. We develop the XAI model using a dataset from the Korean MMORPG, AION, which includes game logs of human players and game bots. More than one classification model has been applied to the dataset to be analyzed by applying interpretable models. This provides us explanations about the game bots' behavior, and the truthfulness of the explanations has been evaluated. Besides, interpretability contributes to minimizing false detection, which imposes unfair restrictions on human players.

Kyung Ho Park, Eunji Park, Huy Kang Kim

Along with the importance of safety, an IDS has become a significant task in the real world. Prior studies proposed various intrusion detection models for the UAV. Past rule-based approaches provided a concrete baseline IDS model, and the machine learning-based method achieved a precise intrusion detection performance on the UAV with supervised learning models. However, previous methods have room for improvement to be implemented in the real world. Prior methods required a large labeling effort on the dataset, and the model could not identify attacks that were not trained before. To jump over these hurdles, we propose an IDS with unsupervised learning. As unsupervised learning does not require labeling, our model let the practitioner not to label every type of attack from the flight data. Moreover, the model can identify an abnormal status of the UAV regardless of the type of attack. We trained an autoencoder with the benign flight data only and checked the model provides a different reconstruction loss at the benign flight and the flight under attack. We discovered that the model produces much higher reconstruction loss with the flight under attack than the benign flight; thus, this reconstruction loss can be utilized to recognize an intrusion to the UAV. With consideration of the computation overhead and the detection performance in the wild, we expect our model can be a concrete and practical baseline IDS on the UAV.