Filter Results
308111 results
We represented a new Bangla dataset with a Hybrid Recurrent Neural Network model which generated Bangla natural language description of images. This dataset achieved by a large number of images with classification and containing natural language process of images. We conducted experiments on our self-made Bangla Natural Language Image to Text (BNLIT) dataset. Our dataset contained 8,743 images. We made this dataset using Bangladesh perspective images. We used one annotation for each image. In our repository, we added two types of pre-processed data which is 224 × 224 and 500 × 375 respectively alongside annotations of full dataset. We also added CNN features file of whole dataset in our repository which is features.pkl.
Data Types:
  • Other
  • Software/Code
  • Dataset
  • Text
  • File Set
The functions used to carry out this work are found in the files provided, "k-Prototypes Clustering" and "clustMixType modified functions". These algorithms carry out the operations of obtaining and manipulating the data matrix, descriptive statistics of the data, determining the best number of clusters, clustering with the k-prototypes method, and statistical validation of the generated clusters with MANOVA. An example is also presented using the Iris database, contained in the R software library, and widely used to exemplify and validate algorithms developed in R language. The functions modified for this work are found in the files "clustMixType modified functions". The modified functions are called in the algorithm of the file "k-Prototypes Clustering", on line 41, by the file "k-Prototypes Clustering.R". The kproto.modif (), clprofiles.modif () and summary.kproto.modif () functions were modified from the kproto (), clprofiles () and summary.kproto () functions, respectively, of the clustMixType package, developed by SZEPANNEK (2018). The dist.binary () function of the ade4 package, developed by DRAY & DUFOUR (2017), was also used in the development of the kproto.modif () function, that now can use a variety of similarity functions. The relationship between the variables is expressed by the squared Euclidean distance, to quantify the distance between numerical variables, and for the nominal variables, the distance can be obtained from a variety of coefficients of similarity. The fviz_cluster.modif () function was modified from the fviz_cluster () function of the factoextra package, developed by KASSAMBARA & MUNDT (2017). REFERENCES: - DRAY, S.; DUFOUR, A.-B. The ade4 Package: Implementing the Duality Diagram for Ecologists. Journal of Statistical Software, v.22, n.4, p.1-20, set. 2017. R Package version 1.7-13. Available at: https://CRAN.R-project.org/package=ade4. https://www.doi.org/10.18637/jss.v022.i04. - KASSAMBARA, A.; MUNDT, F. factoextra: Extract and Visualize the Results of Multivariate Data Analyses. 2017. R Package version 1.0.5. Available at: https://CRAN.R-project.org/package=factoextra. - SZEPANNEK, G. clustMixType: User-Friendly Clustering of Mixed-Type Data in R. The R Journal, v.10, n.2, p.200-208, 2018. R Package version 0.2-1. Available at: https://CRAN.R-project.org/package=clustMixType. https://www.doi.org/10.32614/RJ-2018-048.
Data Types:
  • Software/Code
  • Dataset
  • Text
Energy-Efficient Hybrid Flowshop Scheduling
Data Types:
  • Dataset
  • File Set
This data set contains subject responses for a series of perceptual tasks (dot counting and angle identification), as well as the R code used to analyze them. Please see the included README files for details.
Data Types:
  • Dataset
  • File Set
Preprocessed EEG data, behavioral measures, participants infos and experiment code for the study 'An ecological measure of rapid and automatic face-sex categorization'. Please find additional information in 'READ_ME.txt' file.
Data Types:
  • Other
  • Image
  • Tabular Data
  • Dataset
  • Text
  • File Set
The raw dataset obtained during performance tests of the proposed surrogate-based global optimization algorithm (KASRA) on all testing problems mentioned in the paper is included and stored in some zipped files. The dataset is relevant to the illustrative examples and Tables 1−4 in sections 5 and 6 of the paper. Besides, all MATLAB plots in fig-format are presented.
Data Types:
  • Dataset
  • File Set
This data set contains experimental data on characterization of the nature-inspired geometry for fluid flow through a solid porous regenerator used in active magentocaloric cooling applications. The data is sorted into several groups: 1. Passive data. This data was collected during the passive testing of 6 regenerators, of which 3 were made out of aluminum and 3 out of magnetocaloric material (MCM). The aim of these experiments was to determine the thermal effectiveness of the regenerators and the pressure drop though them. Since the aluminum regenerators had better surface quality and higher permeability than MCM ones, passive tests allowed to compare performance of different flow channel geometries. The passive testing on MCM regenerators was done in order to determine pressure drop and qualitatively evaluate surface roughness of the flow channels. 2. Active data. In this data set, the raw data from active magnetic testing are stored. MCM regenerators were characterized in a small-scale magnetocaloric demonstrator. The aim of this study was to investigate the behavior of the MCM regenerators under thermo-magnetic cycling. Each of the data files in this group contains full experimental data, but only the last 200s from the recorded data is used for the analysis, since the regenerators reached the periodic steady state condition. 3. Adiabatic temperature change measurements. This experiment was done under heating protocol. The sample and the cabinet were cooled to 273 K. The sample was thermally insulated from surroundings. Once the experiment was started, the cabinet temperature was increased to 302 K. The sample was cyclically magnetized and demagnetized until it reached thermal equilibrium with the cabinet. The data set contains recordings of sample temperature when it was magnetized and demagnetized, as well as induction effect, which must be eliminated before data analysis. 4. Microscopy images. There are microscopy images of randomly selected flow channels of each regenerator. 5. VSM and DSC data. These data are available only in Matlab figure format.
Data Types:
  • Other
  • Image
  • Dataset
  • File Set
Basic data
Data Types:
  • Dataset
  • File Set
In active learning, Optimally Balanced Entropy-Based Sampling (OBEBS) method is a selection strategy from unlabelled data. At active zero-shot learning there is not enough information for supervised machine learning method, thus, our sampling strategy was based on unsupervised learning (clustering). The cluster membership likelihoods of the items were essential for the algorithm to connect the clusters and the classes; i.e. to find assignment between them. For best assignment, Hungarian algorithm was used. We developed and implemented adaptive assignment variants of OBEBS method in the software.
Data Types:
  • Software/Code
  • Dataset
  • Text
3x2x2 design factor 1: condition (feature-alone color, feature-alone shape, binding) factor 2: block (two experimental runs/condition/participants) factor 3: group (old vs young participants) In this study, we compared the electrophysiological activity associated with the binding process in young (N=22, mean age=28.5) and old (N=22; mean age=67.4) participants in a change detection task. Analysis of event-related brain potentials (ERP) focused on the differences between feature-alone (color or shape) and feature-conjunction (color-shape) conditions in stimulus encoding. Independently of aging, discrimination ability was significantly attenuated in the binding condition. The effect, however, was more pronounced in old participants. ERP components related to the visual feature detection and processing (posterior N1 and frontal P2) were not modulated in the binding condition. Only in old participants, a late positive ERP component (LPC) was increased signaling the allocation of additional resources.
Data Types:
  • Tabular Data
  • Dataset
  • File Set
5