Identifying and handling data bias within primary healthcare data using synthetic data generators.

Advanced synthetic data generators can simulate data samples that closely resemble sensitive personal datasets while significantly reducing the risk of individual identification. The use of these advanced generators holds enormous potential in the medical field, as it allows for the simulation and sharing of sensitive patient data. This enables the development and rigorous validation of novel AI technologies for accurate diagnosis and efficient disease management. Despite the availability of massive ground truth datasets (such as UK-NHS databases that contain millions of patient records), the risk of biases being carried over to data generators still exists. These biases may arise from the under-representation of specific patient cohorts due to cultural sensitivities within certain communities or standardised data collection procedures. Machine learning models can exhibit bias in various forms, including the under-representation of certain groups in the data. This can lead to missing data and inaccurate correlations and distributions, which may also be reflected in synthetic data. Our paper aims to improve synthetic data generators by introducing probabilistic approaches to first detect difficult-to-predict data samples in ground truth data and then boost them when applying the generator. In addition, we explore strategies to generate synthetic data that can reduce bias and, at the same time, improve the performance of predictive models.

Multi-objective optimization of the coastal groundwater abstraction for striking the balance among conflicts of resource-environment-economy in Longkou City, China.

Rational coastal groundwater planning is of great significance to freshwater supply for sustainable social-economic development, and to environmental protection in case of seawater intrusion (SI). Quantifying the relation among groundwater quality, quantity, and the related social-economic benefits in a coastal region with intense spatio-temporal variation in groundwater abstraction is helpful to the restoration of the coastal aquifer, and the practical policymaking. However, due to the comprehensive reality involving interdisciplinary principles, it is usually difficult to integrate all the main attributes of groundwater resources into a mono-policymaking process, which might lead to biased decisions, producing a series of adverse impacts on the environment and the social economy. This study thereby develops a combined simulation-optimization model (S-O model) in the coastal part of Longkou City, China, for striking the balance among the three main attributes of groundwater, i.e., the groundwater quantity, groundwater quality or its environmental function, and its related economic yield involving the agricultural and industrial sectors. It is seen that the industrial sector contributed over 80% of the economic yield by consuming over 10% of the total groundwater resource, and the massive agricultural use of groundwater was mainly responsible for the SI. The results of the multi-objective optimization provided practical alternative schemes for groundwater abstraction in terms of maximizing economic yield and minimizing SI. Moreover, the decision discrepancy caused by partial management only considering the groundwater quantity and quality would lower the water use efficiency, and then cause unacceptable economic losses for the enterprises and the government. Our research highlights that the interdisciplinary management of groundwater resources based on the S-O model could significantly improve practicability in groundwater policymaking, and provides a typical reference for the other developing regions facing difficulty in groundwater management during coastal urban planning and economic transformation.

Generating high-fidelity synthetic patient data for assessing machine learning healthcare software.

There is a growing demand for the uptake of modern artificial intelligence technologies within healthcare systems. Many of these technologies exploit historical patient health data to build powerful predictive models that can be used to improve diagnosis and understanding of disease. However, there are many issues concerning patient privacy that need to be accounted for in order to enable this data to be better harnessed by all sectors. One approach that could offer a method of circumventing privacy issues is the creation of realistic synthetic data sets that capture as many of the complexities of the original data set (distributions, non-linear relationships, and noise) but that does not actually include any real patient data. While previous research has explored models for generating synthetic data sets, here we explore the integration of resampling, probabilistic graphical modelling, latent variable identification, and outlier analysis for producing realistic synthetic data based on UK primary care patient data. In particular, we focus on handling missingness, complex interactions between variables, and the resulting sensitivity analysis statistics from machine learning classifiers, while quantifying the risks of patient re-identification from synthetic datapoints. We show that, through our approach of integrating outlier analysis with graphical modelling and resampling, we can achieve synthetic data sets that are not significantly different from original ground truth data in terms of feature distributions, feature dependencies, and sensitivity analysis statistics when inferring machine learning classifiers. What is more, the risk of generating synthetic data that is identical or very similar to real patients is shown to be low.

Using a Smart City IoT to Incentivise and Target Shifts in Mobility Behaviour--Is It a Piece of Pie?

Whilst there is an increasing capability to instrument smart cities using fixed and mobile sensors to produce the big data to better understand and manage transportation use, there still exists a wide gap between the sustainability goals of smart cities, e.g., to promote less private car use at peak times, with respect to their ability to more dynamically support individualised shifts in multi-modal transportation use to help achieve such goals. We describe the development of the tripzoom system developed as part of the SUNSET-SUstainable social Network SErvices for Transport-project to research and develop a mobile and fixed traffic sensor system to help facilitate individual mobility shifts. Its main novelty was its ability to use mobile sensors to classify common multiple urban transportation modes, to generate information-rich individual and group mobility profiles and to couple this with the use of a targeted incentivised marketplace to gamify travel. This helps to promote mobility shifts towards achieving sustainability goals. This system was trialled in three European country cities operated as Living Labs over six months. Our main findings were that we were able to accomplish a level of behavioural shifts in travel behaviour. Hence, we have provided a proof-of-concept system that uses positive incentives to change individual travel behaviour.