• Journal of the Korea Society of Computer and Information
• /
• v.27 no.3
• /
• pp.173-182
• /
• 2022

In this paper, we propose a machine learning model that can prevent fraudulent transactions in advance and interpret them using the XAI approach. For the experiment, we collected a real data set of 12,258 mobile phone sales posts from Joonggonara, a major domestic online C2C resale trading platform. Characteristics of the text corresponding to the post body were extracted using Doc2vec, dimensionality was reduced through PCA, and various derived variables were created based on previous research. To mitigate the data imbalance problem in the preprocessing stage, a complex sampling method that combines oversampling and undersampling was applied. Then, various machine learning models were built to detect fraudulent postings. As a result of the analysis, LightGBM showed the best performance compared to other machine learning models. And as a result of SHAP, if the price is unreasonably low compared to the market price and if there is no indication of the transaction area, there was a high probability that it was a fraudulent post. Also, high price, no safe transaction, the more the courier transaction, and the higher the ratio of 0 in the price also led to fraud.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.706-708
• /
• 2023

In this study, we embark on a journey to uncover the essence of emotions by exploring the depths of transfer learning on three pre-trained transformer models. Our quest to classify five emotions culminates in discovering the KLUE (Korean Language Understanding Evaluation)-BERT (Bidirectional Encoder Representations from Transformers) model, which is the most exceptional among its peers. Our analysis of F1 scores attests to its superior learning and generalization abilities on the experimental data. To delve deeper into the mystery behind its success, we employ the powerful SHAP (Shapley Additive Explanations) method to unravel the intricacies of the KLUE-BERT model. The findings of our investigation are presented with a mesmerizing text plot visualization, which serves as a window into the model's soul. This approach enables us to grasp the impact of individual tokens on emotion classification and provides irrefutable, visually appealing evidence to support the predictions of the KLUE-BERT model.

• The Journal of the Korea Contents Association
• /
• v.21 no.12
• /
• pp.13-22
• /
• 2021

The clinical decision support system uses accumulated medical data to apply an AI model learned by machine learning to patient diagnosis and treatment prediction. However, the existing black box-based AI application does not provide a valid reason for the result predicted by the system, so there is a limitation in that it lacks explanation. To compensate for these problems, this paper proposes a system model that applies XAI that can be explained in the development stage of the clinical decision support system. The proposed model can supplement the limitations of the black box by additionally applying a specific XAI technology that can be explained to the existing AI model. To show the application of the proposed model, we present an example of XAI application using LIME and SHAP. Through testing, it is possible to explain how data affects the prediction results of the model from various perspectives. The proposed model has the advantage of increasing the user's trust by presenting a specific reason to the user. In addition, it is expected that the active use of XAI will overcome the limitations of the existing clinical decision support system and enable better diagnosis and decision support.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.8
• /
• pp.333-340
• /
• 2023

Efficient prediction of corporate bankruptcy is an important part of making appropriate lending decisions for financial institutions and reducing loan default rates. In many studies, classification models using artificial intelligence technology have been used. In the financial industry, even if the performance of the new predictive models is excellent, it should be accompanied by an intuitive explanation of the basis on which the result was determined. Recently, the US, EU, and South Korea have commonly presented the right to request explanations of algorithms, so transparency in the use of AI in the financial sector must be secured. In this paper, an artificial intelligence-based interpretable classification prediction model was proposed using corporate bankruptcy data that was open to the outside world. First, data preprocessing, 5-fold cross-validation, etc. were performed, and classification performance was compared through optimization of 10 supervised learning classification models such as logistic regression, SVM, XGBoost, and LightGBM. As a result, LightGBM was confirmed as the best performance model, and SHAP, an explainable artificial intelligence technique, was applied to provide a post-explanation of the bankruptcy prediction process.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.12
• /
• pp.67-77
• /
• 2023

This study represents an innovative research conducted in the smart farm environment, developing a deep learning-based disease and pest detection model and applying it to the Intelligent Internet of Things (IoT) platform to explore new possibilities in the implementation of digital agricultural environments. The core of the research was the integration of the latest ImageNet models such as Pseudo-Labeling, RegNet, EfficientNet, and preprocessing methods to detect various diseases and pests in complex agricultural environments with high accuracy. To this end, ensemble learning techniques were applied to maximize the accuracy and stability of the model, and the model was evaluated using various performance indicators such as mean Average Precision (mAP), precision, recall, accuracy, and box loss. Additionally, the SHAP framework was utilized to gain a deeper understanding of the model's prediction criteria, making the decision-making process more transparent. This analysis provided significant insights into how the model considers various variables to detect diseases and pests.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.1
• /
• pp.25-38
• /
• 2022

With the diversification of payment methods and games, related financial accidents are causing serious problems for users and game companies. Recently, game companies have introduced an Fraud Detection System (FDS) for game payment systems to prevent financial incident. However, FDS is ineffective and cannot provide major evidence based on judgment results, as it requires constant change of detection patterns. In this paper, we analyze abnormal transactions among payment log data of real game companies to generate related features. One of the unsupervised learning models, Autoencoder, was used to build a model to detect abnormal transactions, which resulted in over 85% accuracy. Using X-FDS (Explainable FDS) with XAI-SHAP, we could understand that the variables with the highest explanation for anomaly detection were the amount of transaction, transaction medium, and the age of users. Based on X-FDS, we derive an improved detection model with an accuracy of 94% was finally derived by fine-tuning the importance of features that adversely affect the proposed model.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.18 no.5
• /
• pp.999-1008
• /
• 2023

Early prediction of chronic diseases such as diabetes is an important issue, and improving the accuracy of diabetes prediction is especially important. Various machine learning and deep learning-based methodologies are being introduced for diabetes prediction, but these technologies require large amounts of data for better performance than other methodologies, and the learning cost is high due to complex data models. In this study, we aim to verify the claim that DNN using the pima dataset and k-fold cross-validation reduces the efficiency of diabetes diagnosis models. Machine learning classification methods such as decision trees, SVM, random forests, logistic regression, KNN, and various ensemble techniques were used to determine which algorithm produces the best prediction results. After training and testing all classification models, the proposed system provided the best results on XGBoost classifier with ADASYN method, with accuracy of 81%, F1 coefficient of 0.81, and AUC of 0.84. Additionally, a domain adaptation method was implemented to demonstrate the versatility of the proposed system. An explainable AI approach using the LIME and SHAP frameworks was implemented to understand how the model predicts the final outcome.

• Journal of Intelligence and Information Systems
• /
• v.28 no.1
• /
• pp.263-286
• /
• 2022

This study acquires NBA statistical information for a total of 32 years from 1990 to 2022 using web crawling, observes variables of interest through exploratory data analysis, and generates related derived variables. Unused variables were removed through a purification process on the input data, and correlation analysis, t-test, and ANOVA were performed on the remaining variables. For the variable of interest, the difference in the mean between the groups that advanced to the playoffs and did not advance to the playoffs was tested, and then to compensate for this, the average difference between the three groups (higher/middle/lower) based on ranking was reconfirmed. Of the input data, only this year's season data was used as a test set, and 5-fold cross-validation was performed by dividing the training set and the validation set for model training. The overfitting problem was solved by comparing the cross-validation result and the final analysis result using the test set to confirm that there was no difference in the performance matrix. Because the quality level of the raw data is high and the statistical assumptions are satisfied, most of the models showed good results despite the small data set. This study not only predicts NBA game results or classifies whether or not to advance to the playoffs using machine learning, but also examines whether the variables of interest are included in the major variables with high importance by understanding the importance of input attribute. Through the visualization of SHAP value, it was possible to overcome the limitation that could not be interpreted only with the result of feature importance, and to compensate for the lack of consistency in the importance calculation in the process of entering/removing variables. It was found that a number of variables related to three points and errors classified as subjects of interest in this study were included in the major variables affecting advancing to the playoffs in the NBA. Although this study is similar in that it includes topics such as match results, playoffs, and championship predictions, which have been dealt with in the existing sports data analysis field, and comparatively analyzed several machine learning models for analysis, there is a difference in that the interest features are set in advance and statistically verified, so that it is compared with the machine learning analysis result. Also, it was differentiated from existing studies by presenting explanatory visualization results using SHAP, one of the XAI models.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.31 no.4
• /
• pp.559-571
• /
• 2021

Along with the increase prevalence of computers, the number of malware distributions by attackers to ordinary users has also increased. Research to detect malware continues to this day, and in recent years, research on malware detection and analysis using AI is focused. However, the AI algorithm has a disadvantage that it cannot explain why it detects and classifies malware. XAI techniques have emerged to overcome these limitations of AI and make it practical. With XAI, it is possible to provide a basis for judgment on the final outcome of the AI. In this paper, we conducted malware group classification using XGBoost and Random Forest, and interpreted the results through SHAP. Both classification models showed a high classification accuracy of about 99%, and when comparing the top 20 API features derived through XAI with the main APIs of malware, it was possible to interpret and understand more than a certain level. In the future, based on this, a direct AI reliability improvement study will be conducted.

• Korean Chemical Engineering Research
• /
• v.61 no.4
• /
• pp.523-541
• /
• 2023

As accessibility to 3D printers increases, there is a growing frequency of exposure to chemicals associated with 3D printing. However, research on the toxicity and harmfulness of chemicals generated by 3D printing is insufficient, and the performance of toxicity prediction using in silico techniques is limited due to missing molecular structure data. In this study, quantitative structure-activity relationship (QSAR) model based on data-centric AI approach was developed to predict the toxicity of new 3D printing materials by imputing missing values in molecular descriptors. First, MissForest algorithm was utilized to impute missing values in molecular descriptors of hazardous 3D printing materials. Then, based on four different machine learning models (decision tree, random forest, XGBoost, SVM), a machine learning (ML)-based QSAR model was developed to predict the bioconcentration factor (Log BCF), octanol-air partition coefficient (Log Koa), and partition coefficient (Log P). Furthermore, the reliability of the data-centric QSAR model was validated through the Tree-SHAP (SHapley Additive exPlanations) method, which is one of explainable artificial intelligence (XAI) techniques. The proposed imputation method based on the MissForest enlarged approximately 2.5 times more molecular structure data compared to the existing data. Based on the imputed dataset of molecular descriptor, the developed data-centric QSAR model achieved approximately 73%, 76% and 92% of prediction performance for Log BCF, Log Koa, and Log P, respectively. Lastly, Tree-SHAP analysis demonstrated that the data-centric-based QSAR model achieved high prediction performance for toxicity information by identifying key molecular descriptors highly correlated with toxicity indices. Therefore, the proposed QSAR model based on the data-centric XAI approach can be extended to predict the toxicity of potential pollutants in emerging printing chemicals, chemical process, semiconductor or display process.