31 research outputs found
Inverting multivariate analytic characteristic functions with financial applications
No full textThis dissertation is devoted to multivariate analytic characteristic functions inversion and applications in option pricing, option sensitivities estimation, and some electronic engineering problems. We will show that under certain analytic conditions for characteristic functions, the underlying pdfs and cdfs have exponential tails. The inversion from multivariate characteristic functions to the corresponding pdfs and cdfs can be approximated by the trapezoidal rule conveniently with great accuracy. Monte Carlo methods can be applied for option sensitivity analysis. Under multi-dimensional models, acceptance-rejection method is desirable. Simulating from a distribution without explicit pdf or CDF is then transformed to sampling from an easy-to-simulate distribution. Detailed algorithms are provided and comparisons against classical methods in terms of accuracy and efficiency are included.Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2021-05-01The student, Runqi Hu, accepted the attached license on 2019-04-19 at 10:23.The student, Runqi Hu, submitted this Dissertation for approval on 2019-04-19 at 10:32.This Dissertation was approved for publication on 2019-04-19 at 14:25.DSpace SAF Submission Ingestion Package generated from Vireo submission #13740 on 2019-08-22 at 15:07:23Made available in DSpace on 2019-08-23T20:36:02Z (GMT). No. of bitstreams: 2
HU-DISSERTATION-2019.pdf: 1130914 bytes, checksum: 27289498f24e3d4dc439c5d51cea2ed4 (MD5)
LICENSE.txt: 4205 bytes, checksum: 15761a9a88eae8f7c749d5a6a15b43e4 (MD5)
Previous issue date: 2019-04-19Embargo set by: Seth Robbins for item 112175
Lift date: 2021-08-23T20:36:18Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Only Restriction Lifted for Item 112175 on 2021-08-24T09:15:20Z
Eliminating Catastrophic Overfitting Via Abnormal Adversarial Examples Regularization
No full textSingle-step adversarial training (SSAT) has demonstrated the potential to achieve both efficiency and robustness. However, SSAT suffers from catastrophic overfitting (CO), a phenomenon that leads to a severely distorted classifier, making it vulnerable to multi-step adversarial attacks. In this work, we observe that some adversarial examples generated on the SSAT-trained network exhibit anomalous behaviour, that is, although these training samples are generated by the inner maximization process, their associated loss decreases instead, which we named abnormal adversarial examples (AAEs). Upon further analysis, we discover a close relationship between AAEs and classifier distortion, as both the number and outputs of AAEs undergo a significant variation with the onset of CO. Given this observation, we re-examine the SSAT process and uncover that before the occurrence of CO, the classifier already displayed a slight distortion, indicated by the presence of few AAEs. Furthermore, the classifier directly optimizing these AAEs will accelerate its distortion, and correspondingly, the variation of AAEs will sharply increase as a result. In such a vicious circle, the classifier rapidly becomes highly distorted and manifests as CO within a few iterations. These observations motivate us to eliminate CO by hindering the generation of AAEs. Specifically, we design a novel method, termed Abnormal Adversarial Examples Regularization (AAER), which explicitly regularizes the variation of AAEs to hinder the classifier from becoming distorted. Extensive experiments demonstrate that our method can effectively eliminate CO and further boost adversarial robustness with negligible additional computational overhead.Accepted by NeurIPS 202
Research trends on alphavirus receptors: a bibliometric analysis
No full textBackgroundAlphaviruses are a diverse group of pathogens that have garnered considerable attention due to their impact on human health. By investigating alphavirus receptors, researchers can elucidate viral entry mechanisms and gain important clues for the prevention and treatment of viral diseases. This study presents an in-depth analysis of the research progress made in the field of alphavirus receptors through bibliometric analysis.MethodsThis study encompasses various aspects, including historical development, annual publication trends, author and cited-author analysis, institutional affiliations, global distribution of research contributions, reference analysis with strongest citation bursts, keyword analysis, and a detailed exploration of recent discoveries in alphavirus receptor research.ResultsThe results of this bibliometric analysis highlight key milestones in alphavirus receptor research, demonstrating the progression of knowledge in this field over time. Additionally, the analysis reveals current research hotspots and identifies emerging frontiers, which can guide future investigations and inspire novel therapeutic strategies.ConclusionThis study provides an overview of the state of the art in alphavirus receptor research, consolidating the existing knowledge and paving the way for further advancements. By shedding light on the significant developments and emerging areas of interest, this study serves as a valuable resource for researchers, clinicians, and policymakers engaged in combating alphavirus infections and improving public health
Layer-Aware Analysis of Catastrophic Overfitting: Revealing the Pseudo-Robust Shortcut Dependency
No full textCatastrophic overfitting (CO) presents a significant challenge in single-step adversarial training (AT), manifesting as highly distorted deep neural networks (DNNs) that are vulnerable to multi-step adversarial attacks. However, the underlying factors that lead to the distortion of decision boundaries remain unclear. In this work, we delve into the specific changes within different DNN layers and discover that during CO, the former layers are more susceptible, experiencing earlier and greater distortion, while the latter layers show relative insensitivity. Our analysis further reveals that this increased sensitivity in former layers stems from the formation of pseudo-robust shortcuts, which alone can impeccably defend against single-step adversarial attacks but bypass genuine-robust learning, resulting in distorted decision boundaries. Eliminating these shortcuts can partially restore robustness in DNNs from the CO state, thereby verifying that dependence on them triggers the occurrence of CO. This understanding motivates us to implement adaptive weight perturbations across different layers to hinder the generation of pseudo-robust shortcuts, consequently mitigating CO. Extensive experiments demonstrate that our proposed method, Layer-Aware Adversarial Weight Perturbation (LAP), can effectively prevent CO and further enhance robustness.Accepted by ICML 202
Deep Coupled Integration of CSAC and GNSS for Robust PNT
No full textGlobal navigation satellite systems (GNSS) are the most widely used positioning, navigation, and timing (PNT) technology. However, a GNSS cannot provide effective PNT services in physical blocks, such as in a natural canyon, canyon city, underground, underwater, and indoors. With the development of micro-electromechanical system (MEMS) technology, the chip scale atomic clock (CSAC) gradually matures, and performance is constantly improved. A deep coupled integration of CSAC and GNSS is explored in this thesis to enhance PNT robustness. “Clock coasting” of CSAC provides time synchronized with GNSS and optimizes navigation equations. However, errors of clock coasting increase over time and can be corrected by GNSS time, which is stable but noisy. In this paper, weighted linear optimal estimation algorithm is used for CSAC-aided GNSS, while Kalman filter is used for GNSS-corrected CSAC. Simulations of the model are conducted, and field tests are carried out. Dilution of precision can be improved by integration. Integration is more accurate than traditional GNSS. When only three satellites are visible, the integration still works, whereas the traditional method fails. The deep coupled integration of CSAC and GNSS can improve the accuracy, reliability, and availability of PNT
On the Over-Memorization During Natural, Robust and Catastrophic Overfitting
No full textOverfitting negatively impacts the generalization ability of deep neural networks (DNNs) in both natural and adversarial training. Existing methods struggle to consistently address different types of overfitting, typically designing strategies that focus separately on either natural or adversarial patterns. In this work, we adopt a unified perspective by solely focusing on natural patterns to explore different types of overfitting. Specifically, we examine the memorization effect in DNNs and reveal a shared behaviour termed over-memorization, which impairs their generalization capacity. This behaviour manifests as DNNs suddenly becoming high-confidence in predicting certain training patterns and retaining a persistent memory for them. Furthermore, when DNNs over-memorize an adversarial pattern, they tend to simultaneously exhibit high-confidence prediction for the corresponding natural pattern. These findings motivate us to holistically mitigate different types of overfitting by hindering the DNNs from over-memorization training patterns. To this end, we propose a general framework, Distraction Over-Memorization (DOM), which explicitly prevents over-memorization by either removing or augmenting the high-confidence natural patterns. Extensive experiments demonstrate the effectiveness of our proposed method in mitigating overfitting across various training paradigms.Accepted by ICLR 202
Impact of Changing Trends in Medical Therapy on Transurethral Resection of the Prostate: Two Decades of Change in China
No full textOBJECTIVE To retrospectively assess that over the 2 decades, whether medical therapy has changed indications, patient characteristics, and outcomes in men undergoing transurethral resection of the prostate (1992-2013). METHODS At our institution, medical history of all patients undergoing surgery before 1998, between 2001 and 2003, and between 2011 and 2013 was reviewed. Patient demographics, preoperative clinical profile, clinical management, and operative complications were assessed. RESULTS A total of 1157 patients were enrolled in the study. Mean ages of patients increased from 67.0 to 70.4 years old over the past 2 decades. Furthermore, comorbidities increased significantly as well. Although prostate size and weight of resected tissue increased from 57.3 to 92.3 g and from 24.3 to 36.6 g, the surgical time decreased from 78.21 to 72.29 minutes. From 2011 to 2013, patients undergoing surgery had their catheters remove earlier (from 5.7 to 4.5 days), whose postoperative days in hospital were shorter (from 9.3 to 4.4 days). Although operative complications decreased from 12.3% to 5.7%, especially bleeding, re-operation due to bleeding increased from 0.4% to 2.7%. Moreover, no statistical difference was observed in operative complications between patients with medical therapy and those without medical therapy. CONCLUSION The increasing application of medical therapy resulted in surgical interventions delay. The prostate size was significantly greater, as was the weight of resected tissue. Although patients with medication were older with more comorbidities and larger prostates, surgical technique advancements have benefited them and transurethral resection of the prostate is still considered as a safe and recommendable surgical treatment. (C) 2016 Elsevier Inc.SCI(E)[email protected]
Conceptual Framework and Responding Approach of Treatment Burden of Type 2 Diabetes: a Video Recording-based Analysis
No full textBackground Patients with type 2 diabetes commonly experience a high treatment burden. Currently, both domestic and international researches on the treatment burden of the specific diseases is still in its initial stage. Objective To summarize the conceptual framework of treatment burden related to type 2 diabetes in the Chinese population and explore proactive responding approaches for general practitioners based on video recordings of clinical consultation scenarios. Methods A retrospective analysis of video recordings from general practice training clinics in a standardized training base in Guangdong Province from 2018 to 2019 was conducted by using qualitative research methods such as observation record forms, notes from the fields, encoding-retrieval and thematic analysis, combining with existing conceptual framework of treatment burden. Results A total of 49 video recordings of doctor-patient communication about the treatment burden of type 2 diabetes were extracted from 25 video recordings. All 6 themes of the original conceptual framework were mentioned and 2 new themes (burden of medical information and drug-induced hypoglycemia) were identified by analysis that were mentioned repeatedly. A modified conceptual framework of the treatment burden of type 2 diabetes was finally developed containing 7 observable dimensions including economic, drug, medical management, lifestyle change, healthcare system, time/travel, and medical information burdens and the connotation of subtopics in each dimension. According to the analysis of the response approach, general practitioners who have received training can respond consciously to some of the treatment burdens (medications, medical information, time/travel, lifestyle change) by utilizing the skills of health education, enhanced communication, shared decision-making and motivational interviewing. Conclusion This study constructs a modified conceptual framework of treatment burden for patients with type 2 diabetes. General practitioners can consciously respond to treatment burdens of patients by using effective doctor-patient communication skills, in combination with the identification of conceptual dimensions in clinical practice
Conceptual framework and responding approach of treatment burden of type 2 diabetes mellitus: A video recording-based qualitative analysis
No full textBackground: Patients with type 2 diabetes mellitus (T2DM) frequently face a considerable treatment burden. Research on the specific treatment burden associated with T2DM is still in its early stages both in China and worldwide. Objective: Based on video recordings of clinical diagnostic and therapeutic scenarios, this study aims to summarize the conceptual framework of treatment burden related to T2DM among patients in China Mainland and explore proactive coping strategies of general practitioners (GPs). Methods: The researchers conducted a retrospective qualitative analysis using video records that filmed during 2018-2019 from general practice training clinics of a standardized training base in Guangdong Province of China. This analysis was integrated with the existing conceptual framework of treatment burden, employing qualitative research methods such as observational record forms, field notes, coding extraction and thematic analysis. Results: From 25 video records, 49 instances of doctor-patient communication related to the treatment burden of T2DM were selected and analyzed. All six themes from the original conceptual framework were explored. Additionally, the researchers identified two additional thematic items (medical information burden and drug-induced hypoglycemia). The finding led to the development of a modified conceptual framework for the treatment burden of T2DM, encompassing seven observable dimensions: financial burden, medication burden, medical management burden, lifestyle change burden, healthcare system burden, time/travel burden, and medical information burden. Each sub-theme's definition was expanded accordingly. The consultation process analysis indicated that trained GPs can actively respond to treatment burdens (medication burden, medical information burden, time/travel burden, lifestyle change burden) using skills such as patient education, enhanced communication, shared decision-making and motivational interviewing. Conclusion: The study constructs a modified conceptual framework for the treatment burden of T2DM. By integrating the identification of conceptual dimensions in clinical diagnosis and treatment, it explores how Chinese GPs can use doctor-patient communication skills to proactively address specific treatment burdens of T2DM patients
Correlation of ASA Grade and the Charlson Comorbidity Index With Complications in Patients After Transurethral Resection of Prostate
No full textOBJECTIVE To re-assess the Charlson Comorbidity Index (CCI) and the American Society of Anesthesiologists Physical Status Classification System (ASA grade) as predictive factors of complications after transurethral resection of prostate. METHODS This study retrospectively included and analyzed consecutive patients undergoing transurethral resection of the prostate at Peking University First Hospital between 1992 and 2013. A multi-variate analysis was conducted to evaluate the connection of the ASA and CCI grades with the incidence of complications. RESULTS This paper studied 2326 cases in total. The CCI and ASA grades were significantly correlated, with a Spearman rho of 0.245 (P <.001). No considerable differences among the patient cohorts with different CCI or ASA grades were observed in terms of day of catheter removal, surgical time, and prostate size. In addition, no considerable differences were observed in the different modified Clavien classification system scores of complications among patient cohorts with different grades of CCI. CONCLUSION The majority of complications (86.9%) were of grades I, II, and III, whereas grade IV was less frequent (12.1%), and, after transurethral resection of the prostate, grade V was rare (1%). Males with an ASA grade >= 3 and higher CCI scores were more likely to demonstrate a higher incidence of morbidity than males with a lower grade. However, ASA grades and CCI scores were not independent predictors of complications because of the experience of the surgeon and progress in perioperative management and operative techniques. Therefore, for patients with more comorbidities and higher CCI scores or ASA grades, active surgical intervention is still suggested. (C) 2016 Elsevier Inc.SCI(E)[email protected]
