249 research outputs found
Characterization of Rhizobium 'hedysari' by RFLP analysis of PCR-amplified rDNA and by genomic PCR fingerprinting
No full textEatlog: us asistente digital personal para asistir a mejorar los hábitos saludables
No full textTreballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2017, Director: Petia Radeva[es] En la actualidad, los hábitos saludables relacionados con las dietas están cobrando cada vez más protagonismo. La sociedad está tomando conciencia de lo importante que es llevar una dieta equilibrada para muchos aspectos de su vida. El registro de las comidas es una de las grandes dificultades que sufren muchas de las personas que quieren llevar un seguimiento de su dieta.
EatLog se desarrolla con el objetivo de generar automáticamente el diario de comidas a través de una aplicación para móviles gracias al reconocimiento automático de imágenes. Se implementan algoritmos de reconocimiento basados en la tecnologı́a Deep Learning, concretamente en las redes neuronales convolucionales. Esta tecnologı́a ha permitido desarrollar algoritmos que permiten varios reconocimientos de una imagen, entre los que destaca la comida y la categorı́a. La aplicación crea entonces de forma automática un registro de las comidas conectando con un servidor que contiene los algoritmos de reconocimiento. Se han descargado recetas e información nutricional de ingredientes. De este modo la aplicación genera la información nutricional de las comidas reconocidas. El usuario puede entonces consultar información y establecerse objetivos con respecto a cualquiera de los 26 indicadores nutricionales que se gestionan.
Finalmente, para mejorar los resultados de reconocimiento, se ha procedido a descargar imágenes de comida en alta resolución. Esto ha permitido crear una base de datos de 200 categorı́as de comida que reconoce EatLog (101 correspondientes a la base de datos existente Food101 [9]). Las 99 restantes han sido descargadas en el proyecto, con una media de 800 imágenes por categorı́a
Grab, Pay, and Eat: Semantic Food Detection for Smart Restaurants
No full textThis work was supported in part by TIN2015-66951-C2-1-R, in part by SGR 1219, in part by CERCA Programme/Generalitat de Catalunya, and in part by the NVIDIA Corporation with the donation of the Titan Xp GPU. The work of E. Aguilar was supported by CONICYT Becas Chile. The work of B. Remeseiro was supported by the Spanish Ministry of Economy and Competitiveness under Juan de la Cierva Program (ref. FJCI-2014-21194). The work of M. Bolanos ˜ was supported by an FPU fellowship (ref. FPU15/01347). The work of P. Radeva was supported by ICREA Academia 2014
Characterizing the contribution of dependent features in XAI methods
No full textExplainable Artificial Intelligence (XAI) provides tools to help
understanding how the machine learning models work and reach a specific
outcome. It helps to increase the interpretability of models and makes the
models more trustworthy and transparent. In this context, many XAI methods were
proposed being SHAP and LIME the most popular. However, the proposed methods
assume that used predictors in the machine learning models are independent
which in general is not necessarily true. Such assumption casts shadows on the
robustness of the XAI outcomes such as the list of informative predictors.
Here, we propose a simple, yet useful proxy that modifies the outcome of any
XAI feature ranking method allowing to account for the dependency among the
predictors. The proposed approach has the advantage of being model-agnostic as
well as simple to calculate the impact of each predictor in the model in
presence of collinearity.Comment: 17 pages, 5 table
Predictive analysis of clinical features for HPV status in oropharynx squamous cell carcinoma: A machine learning approach with explainability
No full textBackground and Objective: Oropharynx Squamous Cell Carcinoma (OPSCC) linked to Human Papillomavirus (HPV) exhibits a more favorable prognosis than other squamous cell carcinomas of the upper aerodigestive tract. Finding reliable non-invasive detection methods of this prognostic entity is key to propose appropriate therapeutic decisions. This study aims to provide a comprehensive method based on pre-treatment clinical data for predicting the patient's HPV status over a large OPSCC patient cohort and employing explainability techniques to interpret the significance and effects of the features. Materials and Methods: We employed the RADCURE dataset clinical information to train six Machine Learning algorithms, evaluating them via cross-validation for grid search hyper-parameter tuning and feature selection as well as a final performance measurement on a 20% sample test set. For explainability, SHAP and LIME were used to identify the most relevant relationships and their effect on the predictive model. Furthermore, additional publicly available datasets were scrutinized to compare outcomes and assess the method's generalization across diverse feature sets and populations. Results: The best model yielded an AUC of 0.85, a sensitivity of 0.83, and a specificity of 0.75 over the testing set. The explainability analysis highlighted the remarkable significance of specific clinical attributes, in particular the oropharynx subsite tumor location and the patient's smoking history. The contribution of each variable to the prediction was substantiated by creating a 95% confidence intervals of model coefficients by means of a 10,000 sample bootstrap and by analyzing top contributors across the best-performing models. Conclusions: The combination of specific clinical factors typically collected for OPSCC patients, such as smoking habits and the tumor oropharynx sub-location, along with the ML models hereby presented, can by themselves provide an informed analysis of the HPV status, and of proper use of data science techniques to explain it. Future work should focus on adding other data modalities such as CT scans to enhance performance and to uncover new relations, thus aiding medical practitioners in diagnosing OPSCC more accurately
Telomere length is causally connected to brain MRI image derived phenotypes: A mendelian randomization study
No full textRecent evidence suggests that shorter telomere length (TL) is associated with neuro degenerative diseases and aging related outcomes. The causal association between TL and brain characteristics represented by image derived phenotypes (IDPs) from different magnetic resonance imaging (MRI) modalities remains unclear. Here, we use two-sample Mendelian randomization (MR) to systematically assess the causal relationships between TL and 3,935 brain IDPs. Overall, the MR results suggested that TL was causally associated with 193 IDPs with majority representing diffusion metrics in white matter tracts. 68 IDPs were negatively associated with TL indicating that longer TL causes decreasing in these IDPs, while the other 125 were associated positively (longer TL leads to increased IDPs measures). Among them, ten IDPs have been previously reported as informative biomarkers to estimate brain age. However, the effect direction between TL and IDPs did not reflect the observed direction between aging and IDPs: longer TL was associated with decreases in fractional anisotropy and increases in axial, radial and mean diffusivity. For instance, TL was positively associated with radial diffusivity in the left perihippocampal cingulum tract and with mean diffusivity in right perihippocampal cingulum tract. Our results revealed a causal role of TL on white matter integrity which makes it a valuable factor to be considered when brain age is estimated and investigated
Study of plasma polymer structures to induce composite layers
No full textThis study is designed to investigate the ability of plasma polymer films (PPHMDS), grown from the hexamethyldisiloxane
(HMDS) monomer on stainless steel (SS) and silica glass (SG) substrates, to induce the deposition of composite layers
from a mixture of saturated simulated body fluid (SBF) and detonation nanodiamond (DND) by a biomimetic process.
Results from FTIR and XPS studies showed that the structure of the PPHDMS layers depends on the nature of the
substrate, as well as on the deposition conditions and the influence of the subsequent deposition of the composite layers.
The PPHDMS structure appears to be covalently bonded to SG, compared to those on SS. After their immersion in the
mixture of SBF and DND, the layers grown on the SG_PPHDMS structure shows the existence of phosphate and carbonate
groups. On the SS_PPHMDS, it shows a predominantly carbon enrich deposit, which indicates that the lack of functional
polar groups of the SS_PPHMDS surfaces, and limits the process of precipitation of the SBF ions. The results emphasize
the potential for tailoring a plasma polymer structure PPHMDS, by varying the deposition conditions and substrate, in order
to use them as biocompatible materials
A Perspective on Explainable Artificial Intelligence Methods: SHAP and LIME
No full texteXplainable artificial intelligence (XAI) methods have emerged to convert the black box of machine learning (ML) models into a more digestible form. These methods help to communicate how the model works with the aim of making ML models more transparent and increasing the trust of end-users into their output. SHapley Additive exPlanations (SHAP) and Local Interpretable Model Agnostic Explanation (LIME) are two widely used XAI methods, particularly with tabular data. In this perspective piece, we discuss the way the explainability metrics of these two methods are generated and propose a framework for interpretation of their outputs, highlighting their weaknesses and strengths. Specifically, we discuss their outcomes in terms of model-dependency and in the presence of collinearity among the features, relying on a case study from the biomedical domain (classification of individuals with or without myocardial infarction). The results indicate that SHAP and LIME are highly affected by the adopted ML model and feature collinearity, raising a note of caution on their usage and interpretation
EgoRoutine
No full textThe EgoRoutine is the first dataset for behaviour characterisation in egocentric photo-streams. 100k images collected by 7 different users. @article{talavera2020topic, title={Topic Modelling for Routine Discovery from Egocentric Photo-streams}, author={Talavera, Estefania and Wuerich, Carolin and Petkov, Nicolai and Radeva, Petia}, journal={Pattern Recognition}, pages={107330}, year={2020}, publisher={Elsevier}
EgoRoutine
No full textThe EgoRoutine is the first dataset for behaviour characterisation in egocentric photo-streams. 100k images collected by 7 different users. @article{talavera2020topic, title={Topic Modelling for Routine Discovery from Egocentric Photo-streams}, author={Talavera, Estefania and Wuerich, Carolin and Petkov, Nicolai and Radeva, Petia}, journal={Pattern Recognition}, pages={107330}, year={2020}, publisher={Elsevier}
- …
