Our Work

Prediction of quality and consumers’ preferences is essential task for food producers to improve their market share and reduce any gap in food safety standards. In this paper, we develop a machi…

Prediction of quality and consumers’ preferences is essential task for food producers to improve their market share and reduce any gap in food safety standards. In this paper, we develop a machine learning method to predict yogurt preferences based on the sensory attributes and analysis of samples’ images using image processing texture and color feature extraction techniques. We compare three unsupervised ML feature selection techniques (Principal Component Analysis and Independent Component Analysis and t-distributed Stochastic Neighbour Embedding) with one supervised ML feature selection technique (Linear Discriminant Analysis) in terms of accuracy of classification. Results show the efficiency of the supervised ML feature selection technique over the traditional feature selection techniques.

These files contain the data and the results used in the pilot phase of the tool

Supporting data for the paper: A Robust Geometry-Aware Calibration Method for Non-InvasiveMonopolar EBI Probes: Experimental Validation and Tissue Characterization 

Automatic identification of a script in a given document image facilitates many important applications suchas automatic archiving of multilingual documents, searching online archives of document image…

Automatic identification of a script in a given document image facilitates many important applications suchas automatic archiving of multilingual documents, searching online archives of document images and forthe selection of script specific OCR in a multilingual environment. This paper provides a comparison studyof three dimension reduction techniques, namely partial least squares (PLS), sliced inverse regression (SIR)and principal component analysis (PCA), and evaluates the relative performance of classificationprocedures incorporating those methods. For given script we extracted different features like Gray LevelCo-occurrence Method (GLCM) and Scale invariant feature transform (SIFT) features. The features areextracted globally from a given text block which does not require any complex and reliable segmentation ofthe document image into lines and characters. Extracted features are reduced using various dimensionreduction techniques. The reduced features are fed into Nearest Neighbor classifier. Thus the proposedscheme is efficient and can be used for many practical applications which require processing large volumesof data. The scheme has been tested on 10 Indian scripts and found to be robust in the process of scanningand relatively insensitive to change in font size. This proposed system achieves good classification accuracyon a large testing data set. 

Materials used during the FAIR Clinic: Write an Effective README, delivered as part of the event FAIR Clinics: From Principles to Practice organised by the UMCG Digital Competence Center (DCC) on 8 Oc…

Materials used during the FAIR Clinic: Write an Effective README, delivered as part of the event FAIR Clinics: From Principles to Practice organised by the UMCG Digital Competence Center (DCC) on 8 October 2026.   This workshop focuses on creating project-specific README documentation that improves the understandability and reusability of research outputs. Participants learn how to identify the information that should be documented and develop a README structure tailored to their own research project, dataset, or software.

This repository contains the R analysis code supporting the study “Decomposing mortality risk after allogeneic hematopoietic cell transplantation with a multistate machine learning model.”…

This repository contains the R analysis code supporting the study “Decomposing mortality risk after allogeneic hematopoietic cell transplantation with a multistate machine learning model.”The scripts cover model development and evaluation, comparisons with established prognostic scores, risk phenotype and state occupancy analyses, exploratory comparisons of conditioning regimens, and SHAP analyses. The deposit includes the analysis scripts, settings.R, a fixed predictor bank, a macOS OpenMP build script, and setup instructions.Patient-level data are not included. To run the analyses, users must provide an appropriately authorized dataset named data.xlsx, using the variable names and input coding specified in predictor_bank.xlsx. Please follow readme.docx for software requirements, file placement, and execution instructions.

The EOSC EU Node (EEN) Resource Catalogue is designed to offer discovery and access to the map of resources made available in the EOSC Federation via the EOSC Nodes, including the EEN itself. In parti…

The EOSC EU Node (EEN) Resource Catalogue is designed to offer discovery and access to the map of resources made available in the EOSC Federation via the EOSC Nodes, including the EEN itself. In particular, it aggregates research product profiles (publications, research data, research software) collected from the research product catalogues of EOSC Nodes. The resulting Catalogue is accessible via the EOSC Resource Hub discovery portal (and open APIs), and can be used to discover, access, and monitor EOSC research products across the EOSC Nodes in the Federation. This document outlines the EOSC Resource Catalogue interoperability guidelines that EOSC Node Research Product Catalogue providers should follow to register their catalogues with the EEN Resource Catalogue, making their research products discoverable, visible, and accessible beyond their Node’s boundaries.  

Unpublised and non-submitted article made as part of a course on scientific approach and communication.  The study of nanoparticles is a complex and prolific branch of molecular chemistry, with s…

Unpublised and non-submitted article made as part of a course on scientific approach and communication.  The study of nanoparticles is a complex and prolific branch of molecular chemistry, with significant relevance to research in medicine, electronics, computing, and renewable energy. Carbon quantum dots (CQDs) are a specific type of carbon nanoparticle characterized by high stability, good electrical conductivity, and low toxicity. These structures can confine an electron, which can be excited to produce photoluminescence. Although nanoparticle synthesis may involve expensive reagents and complex processes, the luminescence of CQDs can be demonstrated and modified through a simple experiment using everyday products and household items, making it valuable for educational purposes. Following a theoretical overview of CQDs and their distinctive properties, a synthesis using citric acid (lemonade) and polyethylene glycol (SYSTANE™) is presented. This work also includes a protocol for characterizing the nanoparticles using absorption and emission spectroscopy, as well as nanoparticle tracking analysis (NTA). Finally, the experiment is extended to investigate the quenching of CQD luminescence through controlled additions of Fe²⁺ ions.