Skip to main content

Data Lab for Social Good

The research group will build a critical mass of knowledge and skills in machine learning, forecasting, and management science, empowering social good through informed decision-making.

The research group brings together a diverse and interdisciplinary community of academics and practitioners who are united by their shared dedication to cultivating a strong research expertise in data science, machine learning and forecasting, seamlessly integrated into a set of management science tools, including mathematical modelling, optimisation techniques, systems dynamics, agent-based modelling, and discrete event simulation.

Fueled by an unwavering commitment to advancing social good, the group diligently applies their combined expertise to empower well-informed decision-making and resource optimisation, which in turn leads to tangible and positive societal outcomes


Our aim is to push the boundaries of using data, machine learning, forecasting, and management science to drive social good. Through the development, application, and dissemination of knowledge and skills, we strive to inform and improve decision-making processes that lead to positive societal outcomes. By harnessing the power of these disciplines, we are dedicated to making a meaningful impact and promoting the well-being of communities.


We harness the transformative power of data science and machine learning, seamlessly integrated into a suite of management science tools. By leveraging these capabilities, we transform raw data into actionable insights, empowering decision-makers to drive social good. Our expertise extends to diverse domains, including healthcare operations, humanitarian and global health logistics and supply chain management, agriculture and food supply chain, government, and policy, as well as social sustainability and the circular economy.


Funded projects

Forecasting improvements for better reproductive health and family planning operations in global health supply chains

Dr Bahman Rostami-Tabar, Professor Aris Syntetos, Dr Federico Liberatore and Glenn Milano (USAID)

The aim of this project is to develop innovative machine-learning models for forecasting and subsequently managing contraceptive inventories, and as such improve the performance of reproductive health and family planning supply chains. The models will be empirically routed in and validated (pilot tested) using data from Cote d’Ivoire. The intention is to produce generalisable knowledge that may then be applied to other developing countries.

Collaboration with organisations: United States Agency for International Development (USAID)

Funding: ESRC Wales DTP Collaborative studentships
Duration: September 2023 - September 2026

Understanding and modelling the impact of consumer purchasing behaviour on the global supply chains' decisions in adapting anti-slavery practices

Dr Maryam Lotfi, Dr Bahman Rostami-Tabar, Dr Nicole Koenig-Lewis, Professor Anatoly Zhigljavsky

The project aims to advance knowledge in the field of modern slavery in the supply chain whilst understanding and addressing downstream supply chain challenges, including consumers’ behaviour and attitudes, towards modern slavery and assessing their impact on the upstream supply chains’ decisions on adapting anti-slavery practices.

Collaboration with organisations: UNSEEN
Funding: ESRC Wales DTP Collaborative studentships
Duration: September 2023 - September 2027

Identifying common typologies of harm in forecasting systems

Dr Bahman Rostami-Tabar, Nathaniel Raymond (University of Yale)

This project investigates the typologies of harm and the mechanisms by which they may occur in the forecasting process, which can be generalised, identified, and modified when the life cycle is commonly described. The project produces a catalogue of where forecasting may cause harm, and provide recommendations to address issues of potential harm in the forecasting process.

Funding: Tech Ethics Lab
Duration: September 2022- September 2023

Probabilistic Forecasting of Length of Stay for Inpatient and Community Admissions Due to Mental Health Conditions

Dr Bahman Rostami-Tabar, Siddharth Arora (University of Oxford)

The project investigated the use of probabilistic forecasting model for predicting and understanding the length of stay (LoS) for inpatient and community admissions due to mental health conditions.

Funding: Digital Transformation Innovation Institute
Duration: June 2023 - June 2024

Knowledge distribution and engagement projects

Democratising forecasting

The project entails organising a series of three-day, free-of-charge workshops in low and lower-middle income countries. The primary objective of these workshops is to raise awareness about the significance of forecasting and train university students, academics, and professionals as trainers in the field. The focus is on imparting knowledge and principles of forecasting using R software, which plays a vital role in supporting effective decision-making. By empowering individuals to become trainers themselves, the project aims to foster a wider understanding and application of forecasting techniques within these countries, ultimately contributing to improved decision-making processes.

Democratising Forecasting website

AFRICAST: Empowering Forecasting Excellence across Africa with R and Python

In partnership with Jomo Kenyatta University of Agriculture and Technology ( Department of Statistics) in Kenya, we offer two online training programs on forecasting principles using R and Python. These programs are designed to train 50 learners each, with a total of 100 learners from across Africa. The training will encompass live lectures, hands-on labs, and mentorship, ensuring a well-rounded learning experience. We are grateful to have the support of two exceptional researchers, Jonas Rieger from Karlsruhe Institute of Technology in Germany for the Python training and Mitchell O'Hara-Wild from Monash Business School in Australia for the R training. Additionally, mentors have graciously volunteered to guide and mentor five students each for a year, providing invaluable support and guidance throughout the training journey. This partnership promises to deliver high-quality training and empower learners in the field of forecasting in Africa.

NHS-R community - Principles of Time Series Analysis and Forecasting using R

It is becoming increasingly common for organisations to collect huge amounts of data over time, and existing time series analysis tools are not always suitable to handle the scale, frequency and structure of the data collected. In this workshop, we will look at some new packages and methods that have been developed to handle the analysis of large collections of time series.

MSc Live Projects (Organisation-Collaborative MSc Dissertation Projects)

Every year, our MSc students embark on their summer projects from June to September-October. Each student is provided with dedicated supervision by our academic staff. We offer this valuable resource to organisations free of charge and are delighted to support them with these short-term projects.

Using Hierarchical Forecasting methods to estimate Emergency Department demand in Wales

The aim of the project is to develop forecasts of demand for Emergency attendance that enables comparison of methods (national time series, summed forecasts of local level time series, and hierarchical)

Organisation: NHS Wales Executive, Wales
Supervisor: Dr Bahman Rostami-Tabar
Duration: July – September 2023

Housing and Homelessness Needs of Vulnerable Communities to 2040

The aim of the project is to utilise data science tools to analyse homelessness trends and housing needs. By gathering and analysing homeless data, along with considering factors such as population census and other relevant sources, the project aims to predict the future homeless cohort. Additionally, the project will examine housing needs among specific groups, such as disabled individuals, BAME communities, older people, young people, refugees/asylum seekers, and domestic abuse victims. The insights gained from this analysis will inform decision-making and contribute to addressing these social challenges effectively.

Organisation: Walsall Council, England
Supervisor: Dr Bahman Rostami-Tabar
Duration: July – September 2023

Housing and Homelessness Needs of Vulnerable Communities to 2040

The aim of the project is to improve the accuracy of forecasting outputs by reviewing and enhancing the methods used by four counties. It involves forecasting multiple Maternal, Newborn, and Child Health (MNCH) commodities for the next three years using available data and tools. Additionally, the project aims to develop AI approaches that strengthen forecasting and enable AI-enabled decision-making. The goal is to optimise forecasting accuracy and inform decision-making in the MNCH domain.

Organisation: inSupply Health, Kenya
Supervisor: Dr Bahman Rostami-Tabar
Duration: July – September 2023

The prediction of essential medicines demand at public health facilities in Amhara region, Ethiopia: Using a machine learning approach

The aim of the project is to develop a predictive model that accurately forecasts the demand for essential medicines at public health facilities. The objectives of the project include estimating historical consumption data for essential medicines, developing a robust prediction model for forecasting their demand, and validating the effectiveness of the developed model. By successfully achieving these objectives, the project aims to provide healthcare facilities with a reliable tool for anticipating and meeting the demand for essential medicines, ultimately improving healthcare service delivery and ensuring the availability of vital medications for patients.

Organisation: Ethiopian Pharmaceutical Supply Service, Amhara region, Ethiopia
Supervisor: Bahman Rostami-Tabar
Duration: July – September 2023

Operations Research Identifying opportunities for improvement in the use of a supply chain tracking and planning tool in the supply chain for a fast moving food production and processing company

The aim of the project is to extract insights from unstructured data sets, improve the configuration of the optimisation algorithm, and facilitate more effective system utilisation. By analyzing unstructured data, valuable insights can be derived, enabling informed decision-making. Enhancing the optimisation algorithm's configuration will lead to improved efficiency and accuracy. Additionally, the project aims to develop user-friendly features to aid users in effectively utilising the system and taking actionable steps based on the insights obtained. Overall, the project seeks to maximise the value derived from data, optimise processes, and empower users to make informed decisions.

Organisation: MOY Park, Northern Ireland
Supervisor: Dr Bahman Rostami-Tabar
Duration: June – September 2023


We offer a comprehensive range of training programs specifically designed to empower academics and practitioners with the essential technical and methodological skills to convert raw data into actionable insights.

Our training programs focus on various critical areas, such as machine learning, forecasting, simulation methodologies, reproducibility and open science using Quarto, and proficiency in software tools like R, Python, Matlab, and AnyLogic.

If you seek more specific information about our training programs, please don't hesitate to contact us. We are enthusiastic about providing you with detailed insights to help you make informed decisions for the greater good.


Book launch - Demand Forecasting for Executives and Professionals

5 October 2023, Cardiff Business School



Here you will find the list of schools and research groups we are actively collaborating with.

External organisation partners

Here you will find a list of our actively engaged external organisation partners.

External academic partners

Here you will find a list of our esteemed academic partners with whom we actively collaborate.