Search results for "feature"
showing 10 items of 4091 documents
Dynamic integration of classifiers in the space of principal components
2003
Recent research has shown the integration of multiple classifiers to be one of the most important directions in machine learning and data mining. It was shown that, for an ensemble to be successful, it should consist of accurate and diverse base classifiers. However, it is also important that the integration procedure in the ensemble should properly utilize the ensemble diversity. In this paper, we present an algorithm for the dynamic integration of classifiers in the space of extracted features (FEDIC). It is based on the technique of dynamic integration, in which local accuracy estimates are calculated for each base classifier of an ensemble, in the neighborhood of a new instance to be pr…
Closed-Form Expressions for Global and Local Interpretation of Tsetlin Machines
2021
Tsetlin Machines (TMs) capture patterns using conjunctive clauses in propositional logic, thus facilitating interpretation. However, recent TM-based approaches mainly rely on inspecting the full range of clauses individually. Such inspection does not necessarily scale to complex prediction problems that require a large number of clauses. In this paper, we propose closed-form expressions for understanding why a TM model makes a specific prediction (local interpretability). Additionally, the expressions capture the most important features of the model overall (global interpretability). We further introduce expressions for measuring the importance of feature value ranges for continuous feature…
Embodied sound design
2018
Abstract Embodied sound design is a process of sound creation that involves the designer’s vocal apparatus and gestures. The possibilities of vocal sketching were investigated by means of an art installation. An artist–designer interpreted several vocal self-portraits and rendered the corresponding synthetic sketches by using physics-based and concatenative sound synthesis. Both synthesis techniques afforded a broad range of artificial sound objects, from concrete to abstract, all derived from natural vocalisations. The vocal-to-synthetic transformation process was then automated in SEeD, a tool allowing to set and play interactively with physics- or corpus-based sound models. The voice-dri…
Low-Rank Tucker-2 Model for Multi-Subject fMRI Data Decomposition with Spatial Sparsity Constraint
2022
Tucker decomposition can provide an intuitive summary to understand brain function by decomposing multi-subject fMRI data into a core tensor and multiple factor matrices, and was mostly used to extract functional connectivity patterns across time/subjects using orthogonality constraints. However, these algorithms are unsuitable for extracting common spatial and temporal patterns across subjects due to distinct characteristics such as high-level noise. Motivated by a successful application of Tucker decomposition to image denoising and the intrinsic sparsity of spatial activations in fMRI, we propose a low-rank Tucker-2 model with spatial sparsity constraint to analyze multi-subject fMRI dat…
Zirconium–hafnium and rare earth element signatures discriminating the effect of atmospheric fallout from hydrothermal input in volcanic lake water
2016
The geochemical behaviour of Rare Earth Elements, Zr and Hf was investigated in the thermal waters of Nevado del Ruiz volcano system. A wide range of pH, between 1.0 and 8.8, characterizes these fluids. The acidicwaters are sulphate dominatedwith different Cl/SO4 ratios. The important role of the pH and the ionic complexes for the distribution of REE, Zr a nd Hf in the aqueous phase was evidenced. The pH rules the precipitation of authigenic Fe and Al oxyhydroxides producing changes in REE, Zr, Hf amounts and strong anomalies of Cerium. The precipitation of alunite and jarosite removes LREE from the solution, changing the REE distribution in acidic waters. Y-Ho and Zr-Hf (twin pairs) have a…
Dynamic Economic Load Dispatch using Levenberg Marquardt Algorithm
2018
Abstract Economic Load Dispatch (ELD) is a very important feature of power system network. This work proposes the novel approach which considers the constraint of ramp rate limit (RRL) to solve the ELD problem. It build up the time varying dynamic economic load dispatch in which load dispatching is calculated for each specified time interval, first it is tested with conventional lambda iteration technique and then the outcomes are used to train artificial neural network (ANN) it is based on Levenberg Marquardt algorithm (LMA).As compared with any other ANN method, the Levenberg Marquardt algorithm based dynamic economic load dispatch is more swift and precise. The propose algorithm is teste…
Suggestions to the Reader
1998
Each section of the book consists of two parts that have different goals. The first part, namely the text itself, is systematically developed. It consists of definitions and proven assertions assembled in an organized fashion and with no significant gaps for the reader to fill. All propositions and theorems, unless ready consequences of definitions and previously proven assertions, are proved in detail.
Euro Crisis and plurality: Does the political orientation of newspapers determine the selection and spin of information?
2016
This article studies the impact of right and left moderate political orientation of newspapers on the levels of plurality in the news coverage of the Euro Crisis in 20 newspapers from 10 European countries through a methodology based on Simpson’s D index. The expectation of finding distinct patterns of coverage leading to high levels of plurality was not fully supported and the results have shown that national frames influence levels of overall plurality more than political ideology.
Feature selection using ROC curves on classification problems
2010
Feature Selection (FS) is one of the key stages in classification problems. This paper proposes the use of the area under Receiver Operator Characteristic curves to measure the individual importance of every input as well as a method to discover the variables that yield a statistically significant improvement in the discrimination power of the classification model.
Image-based detection and classification of allergenic pollen
2015
The correct classification of airborne pollen is relevant for medical treatment of allergies, and the regular manual process is costly and time consuming. An automatic processing would increase considerably the potential of pollen counting. Modern computer vision techniques enable the detection of discriminant pollen characteristics. In this thesis, a set of relevant image-based features for the recognition of top allergenic pollen taxa is proposed and analyzed. The foundation of our proposal is the evaluation of groups of features that can properly describe pollen in terms of shape, texture, size and apertures. The features are extracted on typical brightfield microscope images that enable…