Below is a description of my PhD project which I am doing in partnership with the UK Cabinet Office.<\/p>\n\n\n\n
One of the areas that the Joint Data Analysis Centre (JDAC) is interested in is how the needs of the healthcare system will change in the future. This problem can be split into two main areas: forecasting future demand, and simulation the operational aspects of the system.
When looking at demand, one of the ways of doing this is using conventional forecasting techniques to predict how admission numbers will change based on historical data of past admission numbers. This could be done by using forecasting methods such as ARIMA or ETS to capture behaviour such as seasonality. One way to improve these forecasts is to include external information, for example demographic information or the presence of Covid. Being able to forecast population in different areas of the UK and internal migration of different age groups should also help to improve the accuracy of these forecasts further.
When looking at the operational aspect, a main area of interest is the effect that different government policies can have on the performance of hospitals. This can be done by simulating the operation and performance of hospitals under different scenarios.\u00a0 One way to investigate this is combining the demand forecasting with simulation models, for example for hospital flow.
In both forecasting and simulation, it is important to quantify the uncertainty in the results. This is important for JDAC as if they want to use these results to inform government policies, the uncertainty can tell them how certain we are that the forecasted scenarios will happen. In both forecasting and simulation, there are many sources of uncertainty including structural uncertainty, estimation uncertainty and prediction uncertainty. Uncertainty can be measured using different methods including bootstrapping techniques and model averaging. \u00a0One main challenge especially when using results from a forecast in a simulation model is understanding how the uncertainty propagates throughout the whole system.
<\/p>\n\n\n\n
During the MRes year of the STOR-i programme we get the opportunity to research different topics and produce reports and presentations on them. Below you can can find links to the research I did during this year, as well as the research I did as part of my undergraduate degree and during my STOR-i summer internship.<\/p>\n\n\n\n
Item Response Theory (IRT) Models are commonly used in the construction and evaluation of educational tests. They are typically used with categorical data and are probabilistic models for individuals’ responses to a set of items. These models are based on latent factor models which classify individuals into groups based on traits, where everyone with the same trait behaves in the same way. This method can also be used in other scenarios where the group memberships affect the distributions of the latent traits of the individuals. One example of this is in political data, for example using the voting patterns of different senators in the US Senate to try to classify each senator into one of the two main parties. <\/p>\n\n\n\n