• DSAL
  • Publications
  • Studentiship

Data Science Analytics Lab


Overview

The Data Science Analytics Lab (DSAL) is an initiative whose primary objective is to develop relevant statistical theory for specific data types so as to generate required knowledge to support sustainable development.

Vision

To be a leading Centre of excellence for statistical research, innovations and applications


Objectives

  1. To develop statistical theory, algorithms and systems for knowledge extraction and visualization

  2. To generate information on patterns, insights and predictions from diverse data for various applications

  3. To develop a database of relevant statistical information for timely, evidence-based decision making for sustainable development

  4. To develop collaboration across users and producers of statistical information through capacity building.


Activities

  1. Collaborative statistical and applications research at national, regional and international levels

  2. Capacity building in statistical knowledge and the use of statistical packages, such as R programming language

  3. Support industry, policy makers with workable optimal solutions to solve local problems through attachments.


Accomplishments (2023)

  1. DSAL conducted a Data Mining Workshop with specific focus on Online Questionnaire Design for survey data collection on the 20 March 2023. The workshop was presented by Mr. Hamed Al Masharfi, founder of the Data Mining Company, Oman and coordinated by Dr. Iman Al Hasani, Vice Chair, Data Science Analytics Lab, also Assistant Professor at the Department of Statistics. At least 100 participants attended this successful workshop, with each receiving a certificate of attendance. A copy of the presentation in the Arabic language can be downloaded at the URL. datamining

  2. DSAL conducted an Active Learning Workshop with a specific objective of sensitising faculty at the Department of Statistics about the fundamental aspects of active learning in statistics courses. The workshop was presented by Dr. Maryam Al Alawi, Assistant Professor at the Department of Statistics and member of the Active Learning Committee, and coordinated by Dr. Ronald Wesonga, Chair, Data Science Analytics Lab. All academic faculties at the Department of Statistics participated in this interactive workshop. 

activelearning


Contact Us

For details, collaboration or otherwise, Email [wesonga@squ.edu.om].

 

Research Publications

 

 

Year 2023 Journal Publications

  1. Alodat, M. D., & Baklizi, A. (2023). Pak. J. Statist. 2023 Vol. 39 (1), 11-28 Estimation of the Location-Scale Parameters Based on Ranked Set Sampling Combined with Additional Binomial Information. Pak. J. Statist, 39(1), 11-28.

 

Year 2022 Journal Publications

  1. Abdelbasit, K., & Wesonga, R. (2022). A data analytic model to determine regional variation of asthma incidence and other chronic obstructive lung diseases in Oman. Healthcare Analytics, 2, 100074.
  2. Afra Al Manei, Iman Al Hasani and Ronald Wesonga, (2022)Investigating term weighting schemes on the classification performance for the imbalanced text data, Advances and Applications in Statistics 78, 63-82.
  3. Wesonga, R., Bakheit, C., & Ababneh, F. (2022). Cluster modelling of longitudinal disease data: asthma and potential clinical phenotypes. International Journal of Modelling and Simulation, 42(2), 227-239.
  4. Islam, M. M., Wesonga, R., Al Hasani, I., & Al Manei, A. (2022). Prevalence and determinants of mental health issues among university students during COVID-19 pandemic in Oman: An Online Cross-sectional Study. International Research Journal of Public and Environmental Health, 9 (2), 43-54.
  5. Qananwah, Q., Alqudah, A. M., Alodat, M. D., Dagamseh, A., & Hayden, O. (2022). Detecting Cognitive Features of Videos Using EEG Signal. The Computer Journal, 65(1), 105-123.

 

Year 2021 Journal Publications

  1. Wesonga, R., & Abdelbasit, K. (2021). Region as a risk factor for asthma prevalence: statistical evidence from administrative data. Biostatistics & Epidemiology, 5(1), 19-29.
  2. Amadou Sarr, (2021). A generalized Wishart distribution: matrix variate Varma transform, Far East Journal of Theoretical Statistics 63(2), 51-83.
  3. Alodat, M. T., Fayez, M. D. A., & Eidous, O. (2022). On the asymptotic distribution of Matusita's overlapping measure. Communications in Statistics-Theory and Methods, 51(20), 6963-6977.

  4. Al-Hasani, Iman, S.S. (2021). Estimating Eectiveness of Online Geographically-based Advertising Campaigns, Durham theses, Durham University. 
  5. Al Alawi, M. (2021). Spectral clustering and downsampling-based model selection for functional data (Doctoral dissertation, University of Glasgow).
  6. Al Alawi, M., Ray, S., & Gupta, M. A New Functional Data Clustering Technique Based on Spectral Clustering and Downsampling. In Book of Abstracts (p. 103).
  7. Al Shaaibi, M., & Wesonga, R. (2021). Bias dynamics for parameter estimation with missing data mechanisms under logistic model. Journal of Statistics and Management Systems, 24(4), 873-894.
  8. Al-Shukeili, M., & Wesonga, R. (2021). A Novel Minimization Approximation Cost Classification Method to Minimize Misclassification Rate for Dichotomous and Homogeneous Classes. RMS: Research in Mathematics & Statistics, 8(1), 2021627.
  9. Al Hashmi, I. and Sarr, A (2021). Lomax-Pearson VII distribution with application to financial stock returns. Advances and Applications in Statistics, 71(2), 123-140.
  10. Okiring, J., Routledge, I., Epstein, A., Namuganga, J. F., Kamya, E. V., Obeng-Amoako, G. O., ... Wesonga, R. & Nankabirwa, J. I. (2021). Associations between environmental covariates and temporal changes in malaria incidence in high transmission settings of Uganda: a distributed lag nonlinear analysis. BMC public health, 21(1), 1-11.
  11. Bbosa, F. F., Nabukenya, J., Nabende, P., & Wesonga, R. (2021). On the goodness of fit of parametric and non-parametric data mining techniques: the case of malaria incidence thresholds in Uganda. Health and Technology, 11(4), 929-940.

 

 

PhD Student Research


Below are some PhD research studies at the Department of Statistics:


Title & Student's details: Skewed Elliptical Distributions and their Applications: [PhD Student: Iman Al Hashmi - 47910]

Summary:

The main objectives of the proposed research are to construct a new ECD and skewed ECD, and investigate their mathematical properties as well as to use the constructed distributions in applied contexts and compare their performances with the existing results.


Title & Student's details: On multivariate classification and minimization of misclassification rates: [PhD Student: Mubarak Al Shukeili - 29026]

Summary:

The main objective of the study is to develop statistical model aimed to generate optimal separable hyperplanes that minimizes the MCR. Performance of other classification methods and techniques will be studied theoretically. A method based on the MM Principle will be developed. New classification method based on the support vector machine (SVM) will be proposed and validated theoretically and numerically using simulated and real life data.


Title & Student's details: Bias reduction in parameter estimation under missing data conditions: [PhD Student: Muna Al Shaaibi - 46933]

Summary:

The main objective of this study is propose a method of estimating parameters under data missingness condictions. The EM algorithm and MI are popular methods for dealing with missing data which show superiority over the traditional methods. Inspite of these efforts, parameters derived from these methods are still associated with high biases in the estimated parameters. Missing data present various problems. The proposed method will be validated and its performance compared with the current methods using simulation studies as well as real life data.


Title & Student's details:  Parametric sub-distribution hazard model for clustered competing risks: [PhD Student: Noora Al Shanfari - 11906]

Summary: 

The main objective of this study is to analyse clustered competing risk data using parametric approach. In the presence of competing, two different models can be used to analyze competing risk data, the cause-specific hazard model and the subdistribution hazard model. The latter model is used to estimate the effect of the covariates on the cumulative incidence function. However, in medical research, there are applications involving competing risks where individuals may be correlated. in this case, two models can be fit, the frailty model and the marginalized model. Marginal models have a population-averaged interpretation. Few attempts have been made for modeling clustered competing risk data using marginalized models. However, until now no attempts have been made using a parametric approach.


 

MSc Student Research


Title & Student's details:

Hierarchical Support Vector Machine Classifier for Diabetes Control [Nadia Al Habsi, 40448, Fall2023]

Summary:

A crucial indicator of management and medication compliance is the accurate classification
of diabetic patients based on the degree of metabolic control. Endocrinologists create and
constantly monitor set targets for individuals with diabetes in order to reduce the high-risk
effects of metabolic disorders. Nonetheless, accurate classification is preferred to determine
metabolic regulation because of changes in the normal range (4.4 -10 millimoles per liter)
before and after meals. The available metrics do not take into account the significant linked
demographic and related environmental factors, which may have an impact on how
effectively diabetic patient's metabolism is controlled. Additionally, although the support
vector machine (SVM) is one of the most popular statistical method used in classification
through determining optimal linear hyperplane, it is sometimes identified with lower levels of
classification efficiency.

This study has a two-pronged aim, that is to improve the performance of SVM by proposing
hierarchical SVM and to determine the metabolic control among patients with Type 2
Diabetes (T2D) using data from the Oman National Health NCD risk factors 2017 Survey.

 

BSc Student Research


Final Year Project (FYP) SPRING 2023

   

Code/Section

 

Stud ID

 

Student name

Name of Supervisor

Title of project

 

STAT5557/30

123427

Ahmed AL-Qari

 

Dr. Khidir Abdelbasit

Impact Of The COVID-19 Pandemic On The Academic Performance Of Sultan Qaboos University Students

121891

Majdi Al-Jabri

 

STAT5557/20

120560

Marwan Said Al-Omairi

 

Dr. Maryam Al Alawi

Diagnoses Related To Vitamin-D Levels And Other Factors

119882

Yousuf Mubarak Al-Ismaili

 

STAT5557/40

122139

Murshed Saif Al-Sulimani

 

Dr.  Sarr Amadou

Low Density Lipoprotein And Blood Glucose

120865

Ahmed Mansoor Al-Kharousi

 

STAT5557/50

120296

Abdul Jaleel Al-Ruzaiqi

 

Dr. Brahim Benaid

Analysis Of Data Of Most Common Cancers In Oman From 2009 To 2019.

118960

Rashid Khamis Al-Balushi

 

STAT5557/10

122992

Rawnaq Kkalfan Al-Mahrazi

 

Dr. Moh’d Alodat

Investigation Of The Divorce Cases In Oman Between 2016 And 2020

123786

Rifaa Saud Al-Hadhrami

STAT5502/10

125881

Hind Hamed Al-Habsi

Prof. Islam Mazharul

Heteroscedasticity In Regression Analysis: Effects, Sources And Solution

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Miscellaneous Student Research


End Of Course Projects SPRING 2023

  1. STAT2101 Introduction to Statistics Active Learning Collaborative Project

S2101AL002S2101AL001