Experian (United Kingdom)

Abstract Corn is Brazil's second most cultivated crop (21 Mha in 2023, 80% as a second crop), vital for food, feed, and biofuels. This study provides the first country-level mapping of second-crop corn for Brazil, with results for 2023 and the previous land use/cover 20 and 10 years before. Based on these, we estimated CO2 balance from direct land use change (dLUC) and agricultural management practices (MP). Our results show that the share of corn area that was natural vegetation was 12.9% in 2003–2023 and reduced to 3.6% in 2013–2023 period. 20-year dLUC emissions ranged from 0.6 to 0.9 tCO2⋅ha− 1⋅year− 1, depending on the allocation method, but consistently lower than in the literature. Carbon removals from MP ranged from 0.15 to 0.2 tCO2.ha− 1.yr− 1, mitigating 20% of dLUC. dLUC + MP balance for second-crop corn ethanol ranged from 2.3–5.3 gCO2⋅MJ− 1 in 2003–2023, reducing to 0.32–1.5 in the more recent 2013–2023 period.

Although Machine Learning algorithms are solving tasks of ever-increasing complexity, gathering data and building training sets remains an error prone, costly, and difficult problem. However, reusing knowledge from related previouslysolved tasks enables reducing the amount of data required to learn a new task. We here propose a method for reusing a tree-based model learned in a source task with abundant data in a target task with scarce data. We perform an empirical evaluation showing that our method is useful, especially in scenarios where the labels are unavailable in the target task.

Recent studies demonstrate that people are increasingly looking online to assess their health, with reasons varying from personal preferences and beliefs to inability to book a timely appointment with their local medical practice. Records of these activities represent a new source of data about the health of populations, but which is currently unaccounted for by disease surveillance models. This could potentially be useful as evidence of individuals’ perception of bodily changes and self-diagnosis of early symptoms of an emerging disease. We make use of the Experian geodemographic Mosaic dataset in order to extract Type 2 diabetes candidate risk variables and compare their temporal relationships with the search keywords, used to describe early symptoms of the disease on Google. Our results demonstrate that Google Trends can detect early signs of diabetes by monitoring combinations of keywords, associated with searches for hypertension treatment and poor living conditions; Combined search semantics, related to obesity, how to quit smoking and improve living conditions (deprivation) can be also employed, however, may lead to less accurate results.

Adherence to inhaled asthma medication is poor, increasing the likelihood of asthma attacks. Evidence suggests that supported self-management can improve adherence and digital information support systems may be an effective form of communication. However, individuals9 information needs and communication preferences differ. The aim of this project was to develop a framework through which we could identify people in the UK likely to have higher asthma prevalence and risk of exacerbation and describe their communication preferences. Segmentation is the process of dividing the population into groups of similar characteristics. Using hierarchical clustering, a bespoke audience segmentation was created using Mosaic Public Sector consumer classification data combined with health data taken from health surveys and Hospital Episode Statistics. Primary dimensions when forming the clusters were weighted, with the highest weightings including asthma prevalence, control, hospitalisations and comorbidities. 10 clusters were identified showing different profiles according to asthma prevalence and risk factors, and demographics, lifestyle and communication preference. 3 groups were more likely to have asthma that was also uncontrolled. Differences in communication preference, literacy and demographic factors between these groups were present. A further 2 groups who had an average likelihood of asthma prevalence differed in their propensity to manage the condition well. We were able to identify specific profile and communication preferences among socio-economic groups in relation to asthma and asthma management. We will use these to drive uptake of self-management interventions through mass communication channels and evaluate their impact.

This article reminds us that risk needs to be identified before it can be quantified. It points out that risk models are only as good as the people who devised them and the basic assumption needs to be frequently re-examined. Points out the advantages of using well-researched credit scoring based on current — and not just on statutory filings — and of the value of monitoring trends. It reminds the reader of the valuable risk-relevant information held in corporate trees and of the advisability of discovering what kind of corporate companions any given director has promotes awareness. Aware of the unfortunate effects that the counsel of perfection can have on cash-strapped Small and Medium Enterprises (SMEs) in difficult times — with some suggestions as how to avoid the perfect hamstringing the good.

<h3>ABSTRACT</h3> Engagement in media by consumers has emerged as an important area of interest for media providers, advertising agencies, and advertisers. This article empirically investigates a set of dimensions of engagement for three media channels: television, magazines, and the internet. In addition, this study explores the relationship of media engagement to likelihood of purchase as well as the more general construct of advertising receptivity. The results reveal that there is evidence of a strong relationship between engagement in media and purchase intent and ultimately advertising receptivity.

Abstract Household accumulation of debt and arrears on debt, especially among low‐income families, is an extremely topical issue in the UK media and in policy circles. This paper utilises data from the UK's Survey of Low Income Families in order to examine use of credit, and default and arrears, among low‐income families with children. It shows how credit use and accumulation of arrears differ between single parents and couples with children, and also between homeowners and renters. It also briefly examines the persistence of arrears on specific forms of credit using the panel element of the data‐set, now named the Families and Children Survey.

Over the past 20 years there has been massive investment in management information systems, data warehouses, enterprise resource planning systems and analysis technology. However, considering these massive changes, customer decisions are still made using similar techniques. Many customer decisions are made using sequential rules embedded within campaign management, collections, new business and other customer processing systems. But customer decisioning is not just an analytical process based on statistics, since it also requires the input of experts and knowledge about other factors that cannot always be based on past experience — for example, the future actions of competitors or changes in business priorities. However, many organisations have found that sequential rules are often difficult to maintain, since they require specialists who change and manage the code and consequently the systems do not always reflect the many different factors that should be considered in making a decision. This paper addresses the role of customer decision making in organisations and how technology supports this task. It examines whether decisioning technology has moved forward to take advantage of the massive quantities of data and enormous computing power now available, and whether a ‘fuzzy decisioning’ approach would provide better decisions in certain situations. A fuzzy decisioning approach would weight the significant factors in the decisioning process to make a particular outcome more likely. By contrast, a rule-based approach would split groups of customers into different populations based on a logical condition. After reading the paper, readers may also wish to reflect on whether customer decision making is actually a science or an art!

This paper aims to demonstrate the possibility of combining contextual data held at the neighbourhood level and specific data held at the individual level to produce a classification applicable at individual (and household) level. Building on the strength of the MOSAIC postcode classification it presents an individual level segmentation product aimed specifically at the financial services market. The background and methodology are explained in detail and actual examples of the extra discrimination gained are presented.

This paper focuses on the subject of personal and corporate identity in the UK, the ways in which identity can be substantiated, and the methods used by fraudsters to invent identities or copy other people's identities. The setting for this is the Internet, as utilised as a channel of introduction by two market sectors - retail and financial services. The paper considers the availability and suitability of deploying fraud prevention solutions to reduce substantiated fraud and discourage repeat attacks.