Association pattern discovery via theme dictionary models

Author(s):  
Ke Deng ◽  
Zhi Geng ◽  
Jun S. Liu
2022 ◽  
Vol 139 ◽  
pp. 102629
Author(s):  
Haoran Wang ◽  
Haiping Zhang ◽  
Shangjing Jiang ◽  
Guoan Tang ◽  
Xueying Zhang ◽  
...  

Electronics ◽  
2021 ◽  
Vol 10 (6) ◽  
pp. 727
Author(s):  
Moustafa M. Nasralla ◽  
Basiem Al-Shattarat ◽  
Dhafer J. Almakhles ◽  
Abdelhakim Abdelhadi ◽  
Eman S. Abowardah

The literature on engineering education research highlights the relevance of evaluating course learning outcomes (CLOs). However, generic and reliable mechanisms for evaluating CLOs remain challenges. The purpose of this project was to accurately assess the efficacy of the learning and teaching techniques through analysing the CLOs’ performance by using an advanced analytical model (i.e., the Rasch model) in the context of engineering and business education. This model produced an association pattern between the students and the overall achieved CLO performance. The sample in this project comprised students who are enrolled in some nominated engineering and business courses over one academic year at Prince Sultan University, Saudi Arabia. This sample considered several types of assessment, such as direct assessments (e.g., quizzes, assignments, projects, and examination) and indirect assessments (e.g., surveys). The current research illustrates that the Rasch model for measurement can categorise grades according to course expectations and standards in a more accurate manner, thus differentiating students by their extent of educational knowledge. The results from this project will guide the educator to track and monitor the CLOs’ performance, which is identified in every course to estimate the students’ knowledge, skills, and competence levels, which will be collected from the predefined sample by the end of each semester. The Rasch measurement model’s proposed approach can adequately assess the learning outcomes.


2020 ◽  
pp. 1-13
Author(s):  
Mariane da Silva Dias ◽  
Alicia Matijasevich ◽  
Aluísio JD Barros ◽  
Ana Maria B. Menezes ◽  
Bruna Celestino Schneider ◽  
...  

Abstract Objective: We aimed at evaluating the association of maternal pre-pregnancy nutritional status with offspring anthropometry and body composition. We also evaluated whether these associations were modified by gender, diet and physical activity and mediated by birth weight. Design: Birth cohort study. Setting: Waist circumference was measured with an inextensible tape, and fat and lean mass were measured using dual-energy X-ray absorptiometry. Multiple linear regression was used to adjust for possible confounders and allele score of BMI. We carried out mediation analysis using G-formula. Participants: In 1982, 1993 and 2004, all maternity hospitals in Pelotas (South Brazil) were visited daily and all live births whose families lived in the urban area of the city were evaluated. These subjects have been followed up at different ages. Results: Offspring of obese mothers had on average higher BMI, waist circumference and fat mass index than those of normal weight mothers, and these differences were higher among daughters. The magnitudes of the association were similar in the cohorts, except for height, where the association pattern was not clear. In the 1982 cohort, further adjustment for a BMI allele score had no material influence on the magnitude of the associations. Mediation analyses showed that birth weight captured part of this association. Conclusions: Our findings suggest that maternal pre-pregnancy nutritional status is positively associated with offspring BMI and adiposity in offspring. And this association is higher among daughters whose mother was overweight or obese and, birth weight explains part of this association.


2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Andrew K. C. Wong ◽  
Pei-Yuan Zhou ◽  
Zahid A. Butt

AbstractMachine Learning has made impressive advances in many applications akin to human cognition for discernment. However, success has been limited in the areas of relational datasets, particularly for data with low volume, imbalanced groups, and mislabeled cases, with outputs that typically lack transparency and interpretability. The difficulties arise from the subtle overlapping and entanglement of functional and statistical relations at the source level. Hence, we have developed Pattern Discovery and Disentanglement System (PDD), which is able to discover explicit patterns from the data with various sizes, imbalanced groups, and screen out anomalies. We present herein four case studies on biomedical datasets to substantiate the efficacy of PDD. It improves prediction accuracy and facilitates transparent interpretation of discovered knowledge in an explicit representation framework PDD Knowledge Base that links the sources, the patterns, and individual patients. Hence, PDD promises broad and ground-breaking applications in genomic and biomedical machine learning.


Sign in / Sign up

Export Citation Format

Share Document