No description available.

No description available.

As unlabeled data becomes increasingly available, the need for robust data mining techniques increases as well. Clustering is a common data mining tool which seeks to find related, independent patterns in data called clusters. The cluster validation problem addresses the question of how well a given clustering fits the data set. We present CVIC (cluster validation using instance-based confidences) which assigns confidence scores to each individual instance, as opposed to more traditional methods which focus on the clusters themselves. CVIC trains supervised learners to recreate the clustering, and instances are scored based on output from the learners which corresponds to the confidence that the instance was clustered correctly. One consequence of individually validated instances is the ability to direct users to instances in a cluster that are either potentially misclustered or correctly clustered. Instances with low confidences can either be manually inspected or reclustered and instances with high confidences can be automatically labeled. We compare CVIC to three competing methods for assigning confidence scores and show results on CVIC's ability to successfully assign scores that result in higher average precision and recall for detecting misclustered and correctly clustered instances across five clustering algorithms on twenty data sets including handwritten historical image data provided by Ancestry.com.

INTRODUCTION) Modern contraception is a safe and effective clinical service for reducing unintended pregnancy and improving birth spacing for women. Provider counseling is an important factor that may influence women’s decision making, satisfaction, and self-efficacy with contraception use. This study measures women’s perceptions of recent provider interactions and examines the extent to which these perceptions are associated with method satisfaction and confidence in use. We hypothesize that women who perceive more positive interactions are ultimately more satisfied and confident with their contraceptive method choice. METHODS) A cross-sectional survey of adult reproductive-aged women in two southern states (aged 18 to 44 years old) was used for the analysis. The Statewide Survey of Women was conducted in 2017 by NORC at the University of Chicago. Women were asked to rate providers across a series of statements reflecting aspects of patient-centered contraceptive counseling using a 5-point Likert scale. Responses were dichotomized to reflect those with very good/excellent experiences compared to those with less favorable experiences (Poor, Fair, Good). Women were subsequently asked about their level of satisfaction with current contraceptive methods and confidence in use, also using a 5-point Likert scale. These measures were also dichotomized. A chi-squared test for independence and unadjusted logistic regression models were used to examine associations between patient-provider interactions, satisfaction and confidence in contraceptive use. FINDINGS) The survey included 4,281 respondents. The majority of women reported being satisfied with their current contraceptive method (92.6%) and confident in its use (94.9%). Approximately 93% of women who felt respected as a person by their provider also reported being satisfied with their current birth control method, compared to 73% among women who did not feel respected as a person (p<0.001). Among women who felt their provider allowed them to say what mattered to them about their birth control method, 93.5% were satisfied with their current method, compared to 75% among those who did not feel allowed to say what mattered (p<0.001). Women who felt like their provider took their preferences for birth control into consideration were also more satisfied with their current birth control method compared to those who did not feel the same way (93.7% versus 73.5%; p<0.001). Women who felt their provider allowed them to say what mattered and those who felt their provider took their preferences into consideration were also associated with higher rates of confidence in correct use. Women reporting that their provider gave them enough information to make the best decision about their birth control method was also associated with greater confidence in correct use (95.6% versus 87.5%; p=0.007). CONCLUSION) Findings suggest that positive patient-provider interactions are associated with increased satisfaction in current contraceptive method use. Interestingly, women reporting that providers did not give them enough information to make the best decision about birth control methods also reported being less confident in using their current method. Collectively, these findings support existing evidence suggesting that patient-provider interactions are important for patient-centered care and can be used to inform future clinical practice guidelines around contraceptive counseling.

A review of the literature regarding risk management and effective risk communications identified that very few researchers have addressed risk assessment confidence levels when using risk scoring methods. The focus of this research is to develop a definition of risk assessment confidence levels for use in internal project management communications and to evaluate its usefulness. This research defines risk assessment confidence level as "The degree of certainty that the likelihood or consequence score (assigned by the risk assessor) reflects reality." A specific level of confidence is defined based on the types of analyses that were conducted to determine the risk score. A survey method was used to obtain data from a representative sample of risk assessment professionals from industry and academia to measure their opinion on the usefulness of the defined risk assessment confidence levels. The survey consisted of seven questions related to usefulness--four questions addressed the importance of stating confidence levels in risk assessments and three addressed the usableness of the proposed confidence level. Data were collected on the role and experience level of each of the respondents and the survey also included a comment section to obtain additional feedback. The survey generated 364 respondents representing a broad variety of roles associated with decision making and risk management with experience levels from fairly new to experienced risk assessors. The survey data were analyzed by calculating the proportion of respondents who gave negative, neutral and positive responses to the survey questions. An examination of the roles of the survey respondents indicated that no single group was dominant. A non-parametric Kruskal-Wallis test generally failed to reject the hypothesis that the means of the survey response distributions were identical. There was one exception which indicated that there are differences based on role and by inspection of the responses, it appears that decision makers, academics, and others more strongly support the need for confidence level information to reduce the difficulty in making risk based decisions in projects. The survey responses at a confidence level of 95% have a range of errors from 3.84 to 4.97%. Based on the results of the survey, 77 - 83% of those surveyed indicated agreement that knowing the confidence the assessors have in their assessment is important and would improve a management decision. The survey showed that 60 - 86% of the respondents agreed that the confidence levels and their definitions as presented in the survey were usable. The question with the lowest agreement (60%) was related to the way in which the individual levels were defined. The ad-hoc comments provided in the survey were divided into eleven groups based on similarity of the subject of the comment and then examined for common themes. These added additional insight into the results and useful information for future research efforts. This research validates that the use of risk assessment confidence levels is considered to be useful in project risk management. The research also identified several potential areas for future work, including determining the appropriate number of confidence levels that should be defined, refining the definition of the individual confidence level definitions, examining historical perspectives of whether the risk assessments were accurate, examining the concept of shiftability of risk assessments, further research on communication of variability of risk assessments, and research into the usefulness of risk matrices.

No / We present a novel self-supervised framework for monocular image depth learning and confidence estimation. Our framework reduces the amount of ground truth annotation data required for training Convolutional Neural Networks (CNNs), which is often a challenging problem for the fast deployment of CNNs in many computer vision tasks. Our DepthNet adopts a novel fully differential patch-based cost function through the Zero-Mean Normalized Cross Correlation (ZNCC) to take multi-scale patches as matching and learning strategies. This approach greatly increases the accuracy and robustness of the depth learning. Whilst the proposed patch-based cost function naturally provides a 0-to-1 confidence, it is then used to self-supervise the training of a parallel network for confidence map learning and estimation by exploiting the fact that ZNCC is a normalized measure of similarity which can be approximated as the confidence of the depth estimation. Therefore, the proposed corresponding confidence map learning and estimation operate in a self-supervised manner and is a parallel network to the DepthNet. Evaluation on the KITTI depth prediction evaluation dataset and Make3D dataset show that our method outperforms the state-of-the-art results.

No description available.

No description available.

No description available.

No description available.