Now for Something Data Science Ep.3

The 3rd instalment of our data science mini-series showcasing current research projects carried out by the Data Scientist Development Programme at Leeds Institute for Data Analytics (LIDA).

This week’s features 10-minute lightning talks from three data scientists on rock strength, mental health, and missing people.

Each talk will last 10mins + 10mins Q&A.

Agenda

Talk 1 - Triangulating Police and Health Data to Understand Mental Health-Related Demand (Jayita Chakraborty)

Over the past few years, there has been a noticeable increase in police interactions with individuals experiencing mental ill-health. For example, data from 2021-2022 shows that West Yorkshire Police (WYP) flagged mental health concerns in one out of every 25 incidents they attended. Previous research has also shown that triangulating health and policing administrative data can provide unique insights capable of supporting greater understanding of such complex problems. This project seeks to focus on incidents flagged as mental healthrelated by police. Moreover, the project also explores how spatial and temporal patterns of these incidents overlap and intersect with related measures in health data. The health data is based on the Primary Care interactions, selected via appropriate Systematized Nomenclature of Medicine - Clinical Terms (SNOMED-CT) codes. The analysis involves the preparation of raw administrative datasets for each Lower layer Super Output Area (LSOA) based on the Connected Bradford database. Preliminary analysis suggests that the observed misalignment between police demand and mental health need may reflect an underlying under-provision of mental health services. Apart from cleaning, transforming, and verifying large raw administrative datasets to generate research ready datasets for use in projects, this project also employs exploratory data analysis, spatial and longitudinal analyses, and the application of various modelling strategies including both traditional statistical and machine learning approaches where appropriate. the extent to which police demand and mental health need are mis-aligned, which might be indicate of under-provision

Talk 2 - Understanding Missing from Care episodes (Lewis Shaw)

Across England and Wales, a person is reported as missing every 2 minutes on average. Young people make up a majority of missing persons cases, and children in the care system are much more likely to have one or more missing episodes compared to other children. This project used children’s social care data, combined with missing persons reports to explore patterns among children in the care system who go missing, compared with those that don’t. We also conduct analyses on concentrated repeat missing episodes along with the various factors that may impact the risk of children going missing over time.

Talk 3 - Can we estimate rock strength from rock texture alone? (Sadiq Balogun)

The aims of this project are to build a data mining tool to extract useful rock mechanics data from the vast legacy of published papers and open access theses and to use statistically significant datasets to test the hypothesis that rock strength can be predicted using only the grain size and porosity. This could potentially save time and money from lengthy lab tests and thereby accelerate decarbonisation towards net zero. This data mining tool will also serve as a database for researchers to use for other potential geomechanics research.

More about LIDA

Sign up to hear about LIDA events

Find out more about the Data Scientist Development Programme