Task: The City of Riverside wanted to better understand spending within and across departments and detect anomalies in spending. 

Solution:  Students in CS178 build a pipeline in which data was extensively cleaned to extract items from the detailed descriptions provided. A dashboard was built to visualize various metrics about spending patterns. 

Check out the github repo.

Task: The County of Riverside noted that there is a retention issue with their social workers. They wanted to better understand which social workers are leaving and build a ML model to predict those likely to leave. 

Solution: Students in CS178 built a pipeline to clean and create features. They evaluated various models to predict the likelihood of employees 1- leaving, 2-changing department, or 3 - staying. 

Project results expected March 2022. 

Task:  Sarcix Inc. (MapEDU) is a start-up in Riverside that scrapes webpages (curriculum) of medical schools and build a toolkit to allow educators to audit the curriculum with required standards (such as USMLE. 

Solution: The students in CS 179/178 helped a visualization that would showcase which parts of the curriculum was addressed by a given school. The visualization is dynamic and allows educators to add and update modules to see effects of changing the curriculum. 

Check out the github repo. 

Task: A survey was conducted by a cosmetic surgeon to study people’s preference on five facial areas: forehead shadowing, cheek oblique shadowing, cheek horizontal shadowing, eye positioning, and forehead height. The participants were asked to rate each of four pictures in each set 1 being the best looking and 4 the worst. 

Solution: The students in STAT 183 helped in performing the statistical analysis to find out the picture that most people prefer in each area. 

Task: Help-seeking was operationalized as utilizing campus counseling services when needed. A subsample of students that participated in the University of California Undergraduate Experience Survey (UCUES) and completed the modules that addressed mental health issues was used in the study to determine 1) what factors were important in identifying if the students needed help; 2) if the students needed help, but did not seek it, what factors were most important from keeping them from using the service. 

Solution: Logistic regression models were built to identify impact of multiple factors, such as ethnicity, gender, social class, social support, distress, and wellness, etc. on the help-seeking behaviors of college students.

Task: Students that discontinued the Honor Program voluntarily were asked to fill out a survey including seven free response questions: the reasons that motivated them to participate in University Honors, the reasons that prompted them to voluntarily discontinue their participation in University Honors, the actions they took to resolve their concerns prior to deciding to voluntarily discontinue from University Honors, comment on their experience with University faculty and staff, any difficulties they had satisfying University Honors requirement, any feedback on coursework in University Honors, and any suggestion on how University Honors might improve as a program. 

Solution: The students in STAT 183 helped categorize the responses of each question and find out how the responses are associated with honors cohort, gender and ethnicity.

Task:   Assess the effectiveness of a new initiated protocol for insulin, electrolyte and dextrose intravenous infusion in the management of diabetic ketoacidosis. In this study, the conventional ‘one-bag protocol’ of management of diabetic ketoacidosis (DKA) was compared with the ‘two-bag protocol’ which utilizes two bags of fluids, one containing saline and supplemental electrolytes and the other containing the same solution with the addition of 10% dextrose. All patients included in the study met the inclusion criteria and received treatment with either the one-bag system or the two-bag system. 

Solution: Statistical analysis was conducted to compare the base line characteristics, and outcome measures (time to closure of anion gap) as well among two cohorts. An analysis of covariance (ANCOVA) model was developed using time to closure of the anion gap as the dependent variable, and the following admission variables as covariates: patient’s age, weight, BMI, admission pH and anion gap, BHB, blood urea nitrogen (BUN), serum creatinine, serum blood glucose, HgbA1c, and the Charlson Comorbidity Index.