CKChae-Ho Kim

Chae-Ho Kim

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Chae-Ho Kim has a total of 8 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Development and External Validation of a Multi-Component Odorant Detection-Alarm Algorithm Using Hydrogen Sulfide and Ammonia

    by Chae-Ho Kim, Dong-Chul Shin

    This study developed a machine-learning-based detection-alarm algorithm using hydrogen sulfide (H2S) and ammonia (NH3) as core chemical inputs to determine the analytical detection status of 20 odor compounds and evaluated its external applicability at an independent facility. Data collected at 20 min intervals from three odor-emitting facilities were used, with detection defined as a concentration at or above the compound-specific detection limit. Data from a livestock farm and a wastewater treatment facility were used to develop a meteorological–temporal baseline model (M0b) and an alarm model additionally incorporating H2S and NH3 (M4). Model selection, probability calibration, and compound-specific thresholds were determined using nested site–week cross-validation, and the finalized models were applied unchanged to Site C. In the development dataset, mean average precision (AP) increased from 0.362 for M0b to 0.463 for M4 (ΔAP = +0.101), with improvements for 17 of 20 compounds. At Site C, mean AP increased from 0.424 to 0.443 (ΔAP = +0.019), with improvements for 13 compounds. Excluding exploratory methyl isobutyl ketone, mean ΔAP was +0.106 and +0.020, respectively. External evaluation included only four calendar-week groups, yielding a minimum attainable two-sided permutation p-value of 0.125. No compound-specific improvement reached Benjamini–Hochberg-adjusted significance (q < 0.05) in either dataset; the external findings are therefore exploratory. Under fixed-threshold operation, dimethyl sulfide achieved a precision of 0.921 and a false-positive rate of 0.057, but recall was 0.182. These results suggest that H2S and NH3 may provide useful additional information for screening multiple odor compounds, while emphasizing the need for site-specific threshold assessment and further independent validation.

  • Open Access

    Assessment of Machine Learning Models for Chl-a Prediction Using Water Quality Data: Transferability to a Single External Site and Input Variable Configurations

    by Dong-Ho Kim, Chae-Ho Kim, Jun-Hee Shin, Dong-Chul Shin

    Chlorophyll-a (Chl-a) is an indicator of phytoplankton biomass but is not continuously measured at all water-quality monitoring sites. This study evaluated machine learning models for estimating Chl-a from routinely monitored water-quality variables, focusing on temporal validation, evaluation at a single external site, and reduced input availability. Hourly data from 12 South Korean monitoring sites (2021–2025) were used to develop Random Forest (RF), XGBoost, LightGBM, and Extra Trees (ET) models, optimized using Bayesian optimization and five-fold cross-validation. XGBoost achieved the highest random-split internal R2 (0.9219), whereas RF had the highest external-site R2 point estimate at Yangpyeong (0.6928). However, temporal validation using 2025 data yielded substantially lower R2 values (0.3528–0.5255), indicating reduced predictive performance in a subsequent period. SHAP analysis identified dissolved oxygen and pH as the leading contributors to RF predictions. At Yangpyeong, excluding total nitrogen and total phosphorus produced an R2 of 0.6897 versus 0.6928 with all inputs, with no clear difference detected by paired block bootstrap analysis; excluding turbidity reduced performance. These findings suggest the potential for supplementary Chl-a estimation with reduced inputs at Yangpyeong and provide a basis for further transferability assessments at additional monitoring sites.

  • Open Access

    Persistence and Temporally Ordered Prediction of High-Concentration TMA and DMS Episodes in Incomplete Continuous Monitoring Data

    by Chae-Ho Kim, Dong-Chul Shin

    This study quantified the persistence of high-concentration odor episodes in incomplete long-term continuous monitoring data and evaluated, in temporal order, the predictive value of information available at episode onset for subsequent persistence probability. Trimethylamine (TMA) and dimethyl sulfide (DMS), measured at 20 min intervals at three odor-emitting facilities from August 2022 to June 2023, were analyzed. Episodes with confirmed starts were constructed using facility- and compound-specific 90th percentile (q90) thresholds, and right censoring was applied when observations were interrupted. A total of 1991 episodes were included in the primary analysis, and the Kaplan–Meier median duration was 20 min for all six facility–compound combinations. However, S(60) ranged from 15.1% to 25.5%, S(120) from 8.4% to 17.9%, and RMST120 from 38.8 to 49.0 min, indicating that short episodes coexisted with longer-lasting events. When two consecutive non-exceedances were required to confirm episode termination, restricted mean survival time up to 120 min (RMST120) increased to 63.0–71.4 min, demonstrating the influence of episode definition on persistence estimates. In the temporally ordered predictive evaluation, 921 test episodes were used to assess the additional value of onset concentration and meteorological information, and the predictive effects varied by facility, compound, and evaluation horizon. These results demonstrate that the persistence of high-concentration odor episodes can be quantified at the event level even under incomplete continuous monitoring and provide an analytical framework integrating persistence probability estimation with temporally ordered predictive evaluation.

  • Open Access

    OAV-Based Source Characterization of Designated Odor Compounds in Long-Term Continuous Monitoring Data with Severe Missingness

    by Chae-Ho Kim, Dong-Chul Shin, Myeong-Woon Kim

    Odor management requires consideration of both compound concentrations and olfactory impacts determined by odor threshold concentrations. This study evaluated missing-data structures, interpolation performance, odor activity value (OAV)-based odor contributions, and seasonal compositional changes using 20 min monitoring data for 22 designated odor compounds collected from a livestock farm, a wastewater treatment facility, and an anonymized organic waste treatment facility in Eumseong, Republic of Korea (Site C), from 1 August 2022 to 30 June 2023. Because the dataset contained multi-week to multi-month block missing periods, compound-specific interpolation performance was assessed using block holdout validation with BiLSTM, BiGRU, TCN, Transformer, and HistGradientBoostingRegressor models. OAV, summed odor activity value (SOAV), and odor contribution (OC) were calculated primarily from observed concentrations, while fully interpolated series were additionally used for sensitivity comparison to evaluate interpolation-induced bias. Ammonia was excluded from OAV analysis owing to its low valid observation rate. Among the analyzable compounds, odor contribution was dominated by volatile fatty acids and trimethylamine rather than hydrogen sulfide. The livestock farm and wastewater treatment facility showed n-valeric-acid-dominated profiles, whereas Site C showed a mixed profile involving valeric acids and trimethylamine. Seasonal analysis indicated relatively consistent fatty-acid-dominated compositions at the livestock farm and wastewater treatment facility, while Site C showed a possible shift from fatty-acid dominance in warm seasons to trimethylamine dominance in cold seasons. These results provide an OAV-based framework for identifying odor management priorities in long-term continuous odor datasets with severe missingness.


  • Open Access

    Data Structure-Aware Ensemble Modeling for Time-Series Prediction: A Case Study of Sewage Generation

    by Jae-Sang Lee, Chae-Ho Kim, Jun-Hee Shin, Dong-Ho Kim, Dong-Chul Shin

    This study investigates the influence of time-series data structure on model behavior in sewage generation estimation. Annual data from four regions (2017–2023) were analyzed using Random Forest (RF) and Voting Regressor (VR), while structural characteristics were quantified using autocorrelation, mean absolute change rate, and coefficient of variation. The results indicate that model performance varies depending on data structure. Regions with stronger temporal dependency showed more stable model responses, whereas regions with weaker structural consistency exhibited greater variability in outputs. RF tended to be sensitive to localized fluctuations, leading to region-specific variability, while VR maintained more consistent results by reducing individual model bias and variance. These findings demonstrate that model outcomes are influenced not only by algorithmic design but also by the structural properties of the data, emphasizing the importance of incorporating data characteristics into model selection for sewage generation analysis.

  • Open Access

    Clustering as a Prerequisite for Reliable Machine Learning Prediction of Multi-Odor Systems in Wastewater Treatment

    by Su-Chul Yoon, Chae-Ho Kim, Dong-Chul Shin

    Complex odor emissions from wastewater treatment plants consist of multiple volatile compounds that exhibit heterogeneous temporal dynamics and low linear correlations, making accurate prediction and interpretation difficult when analyzed on a single-compound basis. This study investigates whether clustering can serve not only as an exploratory tool but as an essential preprocessing step to enhance machine-learning performance in multi-odor prediction systems. A total of 22 designated odorants were continuously monitored, and their pairwise dependencies were evaluated using Pearson correlation and mutual information. Data-driven clustering was performed through K-means, hierarchical linkage, and principal-component–based latent grouping, and the resulting structures were quantitatively compared with functional-group-based chemical classifications using the consistency ratio and Jaccard similarity index. Cluster validity was further examined using the Silhouette Coefficient, Davies–Bouldin Index, and Calinski–Harabasz Index. The predictive contribution of clustering was verified by training XGBoost regression models on both raw and cluster-structured datasets. The clustered dataset yielded higher predictive accuracy, with increased R2 and reduced MAE and RMSE across most odorants. SHAP analysis further confirmed that clustering improved model interpretability by stabilizing feature contributions and reducing noise-driven importance shifts. The findings demonstrate that clustering is not a supplementary diagnostic tool, but a prerequisite for building reliable, high-performance machine-learning models in complex odor systems. This integrative framework offers a methodological foundation for multi-odor forecasting, source tracking, and next-generation odor management platforms.


  • Open Access

    Machine Learning-Based Estimation of Sewage Treatment Facility Capacity and Design Adequacy: A Case Study in Korea

    by Jae-Sang Lee, Chae-Ho Kim, Dong-Chul Shin

    Accurate estimation of regional sewage generation is essential for designing reliable and resource-efficient treatment facilities. This study developed an ensemble machine-learning framework to estimate annual sewage generation (SG) as the primary output variable, using a combination of demographic, socioeconomic, and environmental indicators across multiple regions in Korea. The proposed Voting Regressor model, trained using data from four highly urbanized regions (Regions A–D), effectively captured nonlinear interactions among variables such as population, business establishments, economically active population, rainfall, and gross regional domestic product (GRDP). A generalization test on an unseen region (Region E) confirmed the model’s robustness and transferability, demonstrating that the framework can reliably adapt to regions with different demographic and industrial characteristics. Comparative analyses showed that the model outperformed both the Random Forest and the conventional per capita unit-load (GU) method in terms of the coefficient of determination (R2), root mean square error (RMSE), and mean absolute error (MAE). SHAP (Shapley Additive Explanations) analysis further identified business establishments and GRDP as the dominant contributors to sewage generation. Moreover, model-based capacity estimations incorporating a 20% safety factor are closely aligned with actual facility capacities, revealing that conventional design standards often apply excessively conservative margins. The findings demonstrate that the proposed machine learning framework can quantitatively assess design adequacy and prevent structural overestimation while maintaining sufficient operational reserves. This data-driven approach provides an interpretable and adaptable foundation for future sewage infrastructure planning and rational capacity design under evolving socioeconomic and environmental conditions.


  • Open Access

    Fabrication of Biochar-Based Marine Buoy Composites from Sargassum horneri: A Case Study in Korea

    by Chae-Ho Kim, Dong-Chul Shin

    The recurrent influx of invasive Sargassum horneri along the coasts of South Korea poses significant ecological and economic challenges, including habitat disruption, aquaculture damage, and shoreline pollution. This study investigates a sustainable valorization pathway by converting SH into functional biochar through slow pyrolysis and utilizing the product as a core material for eco-friendly marine buoys. Biochars were produced at pyrolysis temperatures ranging from 300 °C to 700 °C and characterized for elemental composition, FT-IR spectra, leachability (CODcr), and biodegradability. Higher pyrolysis temperatures resulted in lower H/C and O/C molar ratios, indicating enhanced aromaticity and hydrophobicity. The biochar produced at 700 °C (SFBW-700) exhibited the highest structural and environmental stability, with minimal leachability and resistance to microbial degradation. A composite buoy was fabricated by mixing SFBW-700 with natural binders (beeswax and rosin), forming solid specimens without synthetic polymers or foaming agents. The optimized composition (biochar:beeswax:rosin = 85:10:5) showed excellent performance in density, buoyancy, and impact resistance, while fully meeting the Korean eco-friendly buoy certification criteria. This work presents a circular and scalable approach to mitigating marine macroalgal blooms and replacing plastic-based marine infrastructure with biochar-based eco-friendly composite alternatives. The findings suggest strong potential for the deployment of SH-derived biochar in marine engineering applications.


  • Open Access

    Multivariate Statistical Analysis of Dyeing Wastewater Organic Matter Indicators

    by Youngbeom Cho, June-Shu Ahn, Chae Ho Kim, Dong Chul Shin

    Objectives:Since 2016, TOC (Total Organic Carbon) has replaced COD (Chemical Oxygen Demand) as an organic indicator for effluent wastewater quality standards. However, the distribution of organic substances by process in wastewater treatment facilities is not properly identified, making it difficult to secure stable treated wastewater quality. Therefore, in this study, we identified the correlation between TOC and existing organic matter indicators in raw wastewater, primary treated, secondary treated, and effluent wastewater for dyeing wastewater.Methods:Samples for each process were collected twice a week, a total of 24 times, from a dyeing wastewater treatment plant located in Y-city, Gyeonggi-do, and organic pollutant indicators (TOC, CODCr, CODMn, BOD5) were analyzed. TOC was analyzed by the NPOC (non-purgeable organic carbon) method using TOC-VCHP (Shimadzu, Japan). Using the analysis results, the characteristics of organic pollutants in dyeing wastewater were analyzed. In addition, multivariate statistical analysis was performed using SPSS to analyze correlations between organic pollutant indicators and principal component analysis.Results and Discussion:As a result of multivariate statistical analysis, TOC was inflowed at an average of 574.9 mg/L and treated at 58.2mg/L. In the case of CODCr, CODMn, and BOD5, the inflow was 1,644, 448.9, and 440.7 mg/L and was treated at 98.2, 39.7, and 10.8mg/L. When evaluated based on effluent water quality standards, all of them satisfied the Region III standards, but were discharged at a relatively high level compared to the TOC concentration of sewage treatment plants effluent. As a result of comparing correlations between organic matter indicators through Pearson correlation analysis, the inflow raw water shows a high positive correlation with TOC:TCODCr (r=0.720), TOC:TCODMn (r=0.636), and TOC:TBOD5 (r=0.302) showed low correlation. This is reason to be due to the fact that most organic substances in dyeing wastewater are non-degradable substances and have low biodegradability. As a result of principal component analysis of influent, primary treated, and final treated, three main components each (two for final treated) were extracted, with cumulative contribution rates of 80.1%, 83.2%, and 95.6%.Conclusion:Because the properties of wastewater differ greatly depending on the type of leather and chemicals handled at the dyeing factory, the correlation between influent water was low, but the correlation between treated water and treated water was relatively high. The correlation between processes in wastewater treatment facilities also tended to increase toward later processes. It is believed that the above statistical analysis can be used as basic data for effective organic matter management.