• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.366-371
• /
• 2023

신경망 기반의 검색 모델이 활발히 연구됨에 따라 효과적인 대조학습을 위한 다양한 네거티브 샘플링 방법이 제안되고 있다. 대표적으로, ANN전략은 하드 네거티브 샘플링 방법으로 질문에 대해 검색된 후보 문서들 중에서 정답 문서를 제외한 상위 후보 문서를 네거티브로 사용하여 검색 모델의 성능을 효과적으로 개선시킨다. 하지만 질문에 부착된 정답 문서를 통해 후보 문서를 네거티브로 구분하기 때문에 실제로 정답을 유추할 수 있는 후보 문서임에도 불구하고 네거티브로 분류되어 대조학습을 진행할 수 있다는 문제점이 있다. 이러한 가짜 네거티브 문제(False Negative Problem)는 학습과정에서 검색 모델을 혼란스럽게 하며 성능을 감소시킨다. 본 논문에서는 False Negative Problem를 분석하고 이를 완화시키기 위해 가짜 네거티브 분류기(False Negative Classifier)를 소개한다. 실험은 오픈 도메인 질의 응답 데이터셋인 Natural Question에서 진행되었으며 실제 False Negative를 확인하고 이를 판별하여 기존 성능보다 더 높은 성능을 얻을 수 있음을 보여준다.

• Genomics & Informatics
• /
• v.18 no.2
• /
• pp.20.1-20.4
• /
• 2020

Accessible negative results are relevant for researchers and clinicians not only to limit their search space but also to prevent the costly re-exploration of research hypotheses. However, most biomedical relation extraction datasets do not seek to distinguish between a false and a negative relation among two biomedical entities. Furthermore, datasets created using distant supervision techniques also have some false negative relations that constitute undocumented/ unknown relations (missing from a knowledge base). We propose to improve the distinction between these concepts, by revising a subset of the relations marked as false on the phenotype-gene relations corpus and give the first steps to automatically distinguish between the false (F), negative (N), and unknown (U) results. Our work resulted in a sample of 127 manually annotated FNU relations and a weighted-F1 of 0.5609 for their automatic distinction. This work was developed during the 6th Biomedical Linked Annotation Hackathon (BLAH6).

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.16 no.6
• /
• pp.111-121
• /
• 2006

As Internet lastly grows, network attack techniques are transformed and new attack types are appearing. The existing network-based intrusion detection systems detect well known attack, but the false-positive or false-negative against unknown attack is appearing high. In addition, The existing network-based intrusion detection systems is difficult to real time detection against a large network pack data in the network and to response and recognition against new attack type. Therefore, it requires method to heighten the detection rate about a various large dataset and to reduce the false-positive. In this paper, we propose method to reduce the false-positive using multi-level detection algorithm, that is combine the multidimensional Apriori algorithm and the modified Negative Selection algorithm. And we apply this algorithm in intrusion detection and, to be sure, it has a good performance.

• Analytical Science and Technology
• /
• v.35 no.3
• /
• pp.124-128
• /
• 2022

Blood-sensitive reagents may exhibit false positives or negatives under the influence of substances other than blood. Since these reactions lead to the misinterpretation of blood evidence, it is essential to investigate the possibility of false-positive and -negative reactions of blood-sensitive reagents. Acidic hydrogen peroxide (AHP) is a recently discovered blood-sensitive reagent, and it is not yet known whether it causes false-positive or -negative reactions. To confirm this, 20 µL of blood was placed on metal surfaces, plastic surfaces, paper surfaces, paint surfaces, foods, vegetable oils, detergents, and petroleum hydrocarbons, and then AHP was applied. The blood was observed through an orange filter under a 505-nm light source, and no false-positive or false-negative reactions were observed with any of the substances/materials. However, it was confirmed that polyethylene terephthalate surfaces, polyvinylchloride surfaces, some paint surfaces, and foods exhibit their own photoluminescence under the conditions of blood observation, which interferes with blood observation.

• The Korean Journal of Cytopathology
• /
• v.1 no.1
• /
• pp.51-59
• /
• 1990

Fine needle aspiration biopsy cytology (FNA) for diagnosis of a variety of breast tumors has been proven to be a simple, safe, and cost saving diagnostic methodology with high accuracy. Cytologic specimens from 1,029 fine needle aspirations of the breast during last 3-year period were reviewed and subsequent biopsies from 107 breast lesions were reevaluated for cytohistological correlation. FNA had a sensitivity of 81.6% and a specificity of 98.3%. One oui of 107 cases biopsied revealed a false positive result (0.9%) and the case was due to misinterpretation of apocrine metaplastic cells in necrotic backgound as malignant cells. A false negative rate was 8.4% (9 of 107 cases biopsied). Six of 9 false negative cases were resulted from insufficient aspirates for diagnosis, and remaining three of 9 false negative cases revealed extensive necrosis with no or scanty viable cells on smears. The results indicate that for reducing false positive and false negative rates of FNA, an experienced cytopathologist and a proficient aspirator are of great importance.

• Journal of the Korea Society for Simulation
• /
• v.19 no.4
• /
• pp.139-149
• /
• 2010

Internet was designed for network scalability and best-effort service which makes all hosts connected to Internet to be vulnerable against attack. Many papers have been proposed about attack detection algorithms against the attack using IP spoofing and DoS/DDoS attack. Purpose of DoS/DDoS attack is achieved in short period after the attack begins. Therefore, DoS/DDoS attack should be detected as soon as possible. Attack detection algorithms using false alarm rates consist of the false negative rate and the false positive rate. Moreover, they are important metrics to evaluate the attack detections. In this paper, we analyze the performance of the attack detection algorithms using the impact of false negative rate and false positive rate variation to the normal traffic and the attack traffic by simulations. As the result of this, we find that the number of passed attack packets is in the proportion to the false negative rate and the number of passed normal packets is in the inverse proportion to the false positive rate. We also analyze the limits of attack detection due to the relation between the false negative rate and the false positive rate. Finally, we propose a solution to minimize the limits of attack detection algorithms by defining the network state using the ratio between the number of packets classified as attack packets and the number of packets classified as normal packets. We find the performance of attack detection algorithm is improved by passing the packets classified as attacks.

• Communications for Statistical Applications and Methods
• /
• v.28 no.2
• /
• pp.161-169
• /
• 2021

The two-dimensional confusion matrix used in credit assessment, biostatistics, and many other fields consists of true positive, true negative, false positive, and false negative. Their rates, such as the true positive rate (TPR), true negative rate (TNR), false positive rate, and false negative rate, can be applied to measure its accuracy. In this study, we propose the TPR-TNR plot, a graphical method that can geometrically describe and explain these rates based on the confusion matrix. The proposed TPR-TNR plot consists of two right-angled triangles. We obtain that the TPR and TNR describe the acute angles of right-angled triangles in the plot. These acute angles can be used to determine optimal thresholds corresponding to lots of accuracy measures.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.44 no.4
• /
• pp.154-168
• /
• 2021

COVID-19 has been spreading all around the world, and threatening global health. In this situation, identifying and isolating infected individuals rapidly has been one of the most important measures to contain the epidemic. However, the standard diagnosis procedure with RT-PCR (Reverse Transcriptase Polymerase Chain Reaction) is costly and time-consuming. For this reason, pooled testing for COVID-19 has been proposed from the early stage of the COVID-19 pandemic to reduce the cost and time of identifying the COVID-19 infection. For pooled testing, how many samples are tested in group is the most significant factor to the performance of the test system. When the arrivals of test requirements and the test time are stochastic, batch-service queueing models have been utilized for the analysis of pooled-testing systems. However, most of them do not consider the false-negative test results of pooled testing in their performance analysis. For the COVID-19 RT-PCR test, there is a small but certain possibility of false-negative test results, and the group-test size affects not only the time and cost of pooled testing, but also the false-negative rate of pooled testing, which is a significant concern to public health authorities. In this study, we analyze the performance of COVID-19 pooled-testing systems with false-negative test results. To do this, we first formulate the COVID-19 pooled-testing systems with false negatives as a batch-service queuing model, and then obtain the performance measures such as the expected number of test requirements in the system, the expected number of RP-PCR tests for a test sample, the false-negative group-test rate, and the total cost per unit time, using the queueing analysis. We also present a numerical example to demonstrate the applicability of our analysis, and draw a couple of implications for COVID-19 pooled testing.

• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.716-718
• /
• 2004

The false alarm data in intrusion detection systems are divided into false positive and false negative. The false positive makes bad effects on the performance of intrusion detection system. And the false negative makes bad effects on the efficiency of intrusion detection system. Recently, the most of works have been studied the data mining technique for analysis of alert data. However, the false alarm data not only increase data volume but also change patterns of alert data along the time line. Therefore, we need a tool that can analyze patterns that change characteristics when we look for new patterns. In this paper, we focus on the false positives and present a framework for analysis of false alarm pattern from the alert data. In this work, we also apply incremental data mining techniques to analyze patterns of false alarms among alert data that are incremental over the time. Finally, we achieved flexibility by using dynamic support threshold, because the volume of alert data as well as included false alarms increases irregular.

• The Korean Journal of Cytopathology
• /
• v.7 no.2
• /
• pp.111-121
• /
• 1996

Fine needle aspiration cytology(FNAC) is preferred because of simplicity, safety, and reliability in the evaluation of patients with thyroid nodule or hyperplasia. However, there are a few limitations such as false-negative or false-positive cases and non-diagnostic material. To evaluate the usefulness of FNAC in thyroid lesions, we reviewed 704 FNAC cases of thyroid nodules from 1988 to 1994 at Soonchunhyang University Hospital. The results are as follows. 1. Among 704 FNAC cases of thyroid gland, 571(81.1%) cases were benign, 12(1.7%) were suspicious, 71(10.1%) were malignancy, and 50(7.1%) were material insufficiency. The cytologic diagnoses of the benign lesions included 168 cases of follicular neoplasm, 139 cases of adenomatous goiter, 162 cases of follicular lesion such as follicular neoplasm or adenomatous goiter, 61 cases of Hashimoto's thyroiditis, 13 cases of subacute thyroiditis, and 28 cases of colloidal nodule or benign nodule. The malignant lesions included 68 cases of papillary carcinona, two medullary carcinomas and a case of metastatic colon cancer. 2. The average number of cytologic smear slides was $4.12{\pm}1.81$ in material insufficiency and $5.63{\pm}1.79$ in diagnostic cases. This difference was statistically significant(p<0.00001). 3. Histological assessment of 150 cases revealed 2 false negative and 1 false positive cases. The false negative cases were a case of marked sclerosis in papillary carcinoma and an occult case of papillary carcinoma. The false positive case resulted from pseudo-ground glass nuclei due to marked dry artifact. 4. Comparison between the FNAC and the histologic diagnosis revealed that FNAC had a sensitivity of 93.5%, a specificity of 99.2%, a false negative rate of 6.6%, a false positive rate of 0.8%, and an overall diagnostic accuracy of 98.0%. Therefore, FNAC of thyroid gland is a very reliable diagnostic method with excellent accuracy rate.