Jim Smith's Outputs (136)

Analyst-driven XAI for time series forecasting: Analytics for telecoms maintenance (2024)
Presentation / Conference Contribution

Time series forecasting facilitates real-time anomaly detection in telecom networks, predicting events that disrupt security and service. Current research efforts have been found to focus on new forecasting libraries, more rigorous data cleaning meth... Read More about Analyst-driven XAI for time series forecasting: Analytics for telecoms maintenance.

The inadvertently revealing statistic: A systemic gap in statistical training? (2024)
Journal Article

While concerns around data privacy are well-known, there's a lack of awareness and training when it comes to the confidentiality risk of published statistics, argue Ben Derrick, Elizabeth Green, Felix Ritchie, Jim Smith, Paul White

Machine learning models in trusted research environments - Understanding operational risks (2023)
Journal Article

IntroductionTrusted research environments (TREs) provide secure access to very sensitive data for research. All TREs operate manual checks on outputs to ensure there is no residual disclosure risk. Machine learning (ML) models require very large amou... Read More about Machine learning models in trusted research environments - Understanding operational risks.

A decision-making architecture for human-robot collaboration: Model transferability (2023)
Presentation / Conference Contribution

In this paper, we aim to demonstrate the potential for wider-ranging capabilities and ease of transferability of our recently developed decision-making architecture for human-robot collaboration. To this end, a somewhat related but different applicat... Read More about A decision-making architecture for human-robot collaboration: Model transferability.

SACRO guide to statistical output checking (2023)
Other

This guide for output SDC is the first report from the SACRO project. It covers, theory of output SDC, including the new statbarns model, practicalities, operational considerations, and FAQs for output checking teams.

Disclosure control issues in complex medical data (2023)
Presentation / Conference Contribution

The covid19 pandemic assisted the acceleration of routine access to medical records for research. In the UK platforms including OpenSafely and NHSDigital, alongside emerging hospital trust based Trusted Research Environments (TREs), demonstrate the u... Read More about Disclosure control issues in complex medical data.

SACRO: Semi-Automated Checking Of Research Outputs (2023)
Presentation / Conference Contribution

Output checking can require significant resources, acting as a barrier to scaling up the research use of confidential data. We report on a project, SACRO, that is developing a general-purpose, semi-automatic output checking systems that works across... Read More about SACRO: Semi-Automated Checking Of Research Outputs.

Disclosure control of machine learning models from trusted research environments (TRE): New challenges and opportunities (2023)
Journal Article

Introduction: Artificial intelligence (AI) applications in healthcare and medicine have increased in recent years. To enable access to personal data, Trusted Research Environments (TREs) (otherwise known as Safe Havens) provide safe and secure enviro... Read More about Disclosure control of machine learning models from trusted research environments (TRE): New challenges and opportunities.

A novel mirror neuron inspired decision-making architecture for human–robot interaction (2023)
Journal Article

Inspired by the role of mirror neurons and the importance of predictions in joint action, a novel decision-making structure is proposed, designed and tested for both individual and dyadic action. The structure comprises models representing individual... Read More about A novel mirror neuron inspired decision-making architecture for human–robot interaction.

Inter-annotator agreement using the Conversation Analysis Modelling Schema, for dialogue (2022)
Journal Article

We present the Conversation Analysis Modeling Schema (CAMS), a novel dialogue labeling schema that combines the Conversation Analysis concept of Adjacency Pairs, with Dialogue Acts. The aim is to capture both the semantic and syntactic structure of d... Read More about Inter-annotator agreement using the Conversation Analysis Modelling Schema, for dialogue.

Sentence encoding for dialogue act classification (2021)
Journal Article

In this study, we investigate the process of generating single-sentence representations for the purpose of Dialogue Act (DA) classification, including several aspects of text pre-processing and input representation which are often overlooked or under... Read More about Sentence encoding for dialogue act classification.

Automatic Checking of Research Outputs (ACRO): A tool for dynamic disclosure checks (2021)
Journal Article

This paper discusses the issues surrounding the creation of an automatic tool to reduce the burden of output checking in research environments. It describes ACRO (Automatic Checking of Research Outputs), a Stata tool written as a proof-of-concept, an... Read More about Automatic Checking of Research Outputs (ACRO): A tool for dynamic disclosure checks.

Statistical disclosure controls for machine learning models (2021)
Presentation / Conference Contribution

Artificial Intelligence (AI) models are trained on large datasets. Where the training data is sensitive, the data holders need to consider risks posed by access to the training data and risks posed by the models that are released. The first problem c... Read More about Statistical disclosure controls for machine learning models.

Protein structured reservoir computing for spike-based pattern recognition (2021)
Journal Article

Nowadays we witness a miniaturisation trend in the semiconductor industry backed up by groundbreaking discoveries and designs in nanoscale characterisation and fabrication. To facilitate the trend and produce ever smaller, faster and cheaper computin... Read More about Protein structured reservoir computing for spike-based pattern recognition.

Using active learning to understand the videoconference experience: A case study (2020)
Presentation / Conference Contribution

Videoconferencing is becoming ubiquitous, especially so during the COVID-19 pandemic. However, user experience of a videoconference call can be variable. To better understand and classify the performance of videoconference call systems, this paper re... Read More about Using active learning to understand the videoconference experience: A case study.

Understanding output checking (2020)
Report

This report for Eurostat (Methodology) considers the conceptual and practical issues that need to be addressed in designing and implementing automatic disclosure control checking for statistical research outputs.
The report covers
- The basic theo... Read More about Understanding output checking.

Visual analytics for collaborative human-machine confidence in human-centric active learning tasks (2019)
Journal Article

Active machine learning is a human-centric paradigm that leverages a small labelled dataset to build an initial weak classifier, that can then be improved over time through human-machine collaboration. As new unlabelled samples are observed, the mach... Read More about Visual analytics for collaborative human-machine confidence in human-centric active learning tasks.

Evolutionary n-level hypergraph partitioning with adaptive coarsening (2019)
Journal Article

Hypergraph partitioning is an NP-hard problem that occurs in many computer science applications where it is necessary to reduce large problems into a number of smaller, computationally tractable sub-problems. Current techniques use a multilevel appro... Read More about Evolutionary n-level hypergraph partitioning with adaptive coarsening.

Confidentiality and linked data (2018)
Book Chapter

This chapter considers the confidentiality issues around linked data. It notes that the use and availability of secondary (adminstrative or social media) data, allied to powerful processing and machine learning techniques, in theory means that re-ide... Read More about Confidentiality and linked data.

E-assessment of computer programming (2018)
Presentation / Conference Contribution

This paper demonstrates how we have used Dewis, an algorithmic open source e-assessment system, to automatically assess programming skills, in particular, in the C programming language. Teaching and assessing programming skills is challenging; prior... Read More about E-assessment of computer programming.