Find research datasets worth reusing

opencc-by-4.0Oct 2021View details →

Multiplexed protein profiling reveals spatial subcellular signaling networks

<ul> <li>Supplementary videos for "Multiplexed protein profiling reveals spatial subcellular signaling networks"</li> <li>The videos were made using Volocity 3D visualization software. </li> </ul>

opencc-by-4.0Oct 2021View details →

Data for "Stable climate simulations using a realistic GCM with neural network parameterizations for atmospheric moist physics and radiation processes"

This is sampling data of "Stable climate simulations using a realistic GCM with neural network parameterizations for atmospheric moist physics and radiation processes". 'qv_nn_in' for the large scale specific humidity, [kg/kg] 'T_nn_in' for the large scale temperature, [K] 'dqvls_nn_in' for the large scale moisture advection, [kg/kg/s] 'dTls_nn_in' for the large scale moisture advection, [K/s] 'qtend_check' for the moistening rate by CRM, [kg/kg/s] 'stend_check' for the heating rate by CRM, [K/s] 'SOLS' for direct shorwave solar radiation down to surface, [W/m2] 'SOLSD' for diffusive shortwave solar radiation down to surface, [W/m2] 'SOLL' for direct near infrared solar radiation down to surface, [W/m2] 'SOLLD' for diffusive near infrared solar radiation down to surface, [W/m2] 'SOLIN' for insolation at model top, [W/m2] 'FSNS' for net shortwave radiation at model surface, [W/m2] 'FSNT' for net shortwave radiation at model top, [W/m2] 'FLNS' for net longwave radiation at model surface, [W/m2] 'FLNT' for net longwave radiation at model top, [W/m2] 'SPPS' for surface pressure, [Pa] To download the full dataset of the SPCAM simulation in 1998. Please click the dropbox link: https://www.dropbox.com/s/p841v1tw00rokdy/SPCAM_VAR_1998.tar.gz?dl=0

opencc-by-4.0Nov 2021View details →

Dataset for "Face Detection in Untrained Deep Neural Networks"

Dataset for "Face Detection in Untrained Deep Neural Networks" Seungdae Baek, Min Song, Jaeson Jang, Gwangsu Kim, and Se-Bum Paik* *Contact: <a href="mailto:sbpaik@kaist.ac.kr">sbpaik@kaist.ac.kr</a>   To run demo codes for "Face Detection in Untrained Deep Neural Networks", please download files below.   1. Stimulus.zip - IMG_cntr_210521.mat : A low-level feature-controlled stimulus set was used to find units that responded selectively to face images (Stigliani, 2015). Specifically, 260 images were prepared for each class (face, hand, horn, flower, chair, and scrambled face). - IMG_var_pos/size/rot_210521.mat : To investigate the invariance of face-selective units to face images of various sizes, positions, and rotation angles, the image set (Stigliani, 2015) was generated after modifying the size, position, and rotation angle of the faces and other objects in the low-level feature-controlled stimulus set. - IMG_var_view_210106.mat : This set was used to find units that invariantly responded to face images of different viewpoints. This dataset consists of five angle-based viewpoint classes (-90°, -45°, 0°, 45°, 90°) with ten different faces obtained from (Gourier 2004)   2. Data.zip - Data_PFI_XDream/RevCorr_ClsUnit.mat : The obtained preferred feature images, using reverse-correlation method and X-Dream, of units selective to each object class. - Data_SVM_NeuronType.mat : Simulated face detection performance using distinct types of selective units in Conv5 and using the shuffled responses of face-selective units in untrained networks (a. all selective units, b. units selective to non-face classes (nonface-selective), c. face-selective units, d. units selective to none of these classes (non-selective)). - Data_SVM_invariance.mat : Face detection performance with variation of the low-level features.  - PretrainedNet : Sample networks was untrained and trained with three types of image sets (a. face-reduced ImageNet, b. original ImageNet, c. original ImageNet with added face images (Stigliani, 2015)) (number of networks = 3).    - Net_Untrained : untrained AlexNets   - Net_FD_ImageNet : AlexNets trained to face-reduced ImageNet   - Net_ImageNet : AlexNets trained to original ImageNet   - Net_ImageNetwFace : AlexNets trained to original ImageNet with added face images - Data_Trained.mat : Result summary of four types of networks above (number of networks = 10).

opencc-by-4.0Nov 2021View details →

Heme oxygenase pfam14518 sequence similarity network

Sequence similarity network generated from pfam14518 using the EFI-GNT webtool. Each node is representative of sequences that are 80% identical.

Data in support of: Species-specific interactions in an avian-bryophyte dispersal network

Animal dispersal of plant propagules fundamentally alters the success of dispersal events, and thus shapes plant community composition through time. While this is well-documented in seed plants, spore-bearing plants have received little attention with regard to this phenomenon. Birds are particularly attractive as a potential bryophyte dispersal vector given their highly motile nature as well as their association with bryophytes when foraging and building nests. Despite this, species-specific dispersal relationships between birds and bryophytes have never been examined. We captured birds in Gifford Pinchot National Forest in the Pacific Northwest of the United States to sample their legs and tails for bryophyte spores. We found 24 bryophyte species across 34 species of bird. We examined the level of specialization 1) within the overall interaction network to assess community-level patterns and 2) at the plant species level to determine the effect of bird behavioral type on the plant-animal interaction. Our results suggest that associations within the network are more constrained (specialized) than expected by chance. Additionally, we found that avian foraging guild impacted the variety of bryophytes found on an individual bird. Foliage gleaners and ground foragers had particularly specialized associations within the overall disperser-bryophyte network. Our findings suggest that diffuse bird-bryophyte dispersal networks are likely to be common in habitats where birds readily encounter bryophytes and that further work aimed at understanding individual bird-bryophyte species relationships may prove valuable in determining nuance within this newly described dispersal mechanism.

Patch size distribution affects species invasion dynamics in dendritic networks

<div class="page"> <div class="section"> <div class="layoutArea"> <div class="column"> Biological invasions are globally affecting ecosystems, causing local species loss and altering ecosystem functioning. Understanding how such biological invasions occur and succeed is thus of high priority. Both local properties and the spatial network structure have been shown to be determinants of invasion success, and the identification of spatial invasion hubs directly promoting invasion dynamics is gaining attention. Spatial dynamics, however, could also indirectly alter invasion success by shaping pre- invasion local community structure: in many ecosystems, such as riverine networks, regional properties such as patch size distribution are known drivers of local community structures, which themselves may affect the establishment success of invading species. Using microcosm experiments in dendritic networks, we disentangled how inherent patch size distribution and dispersal along specific network topologies shaped local resident communities, and, subsequently, affected the establishment success of invading species. After controlling for regional-scale effects of connectivity on pre-invasion diversity, we find that patch size distributions independently shaped pre-invasion community diversity and invasion success, with no direct effect of pre-invasion diversity on invasion success. Our results suggest that 1) landscape configuration plays an underestimated role in invasion success and that 2) invasion success should follow predictable landscape-scale patterns in riverine networks given non-random patch-size distribution. </div> </div> </div> </div>

opencc-zeroDec 2021View details →

Chimpanzee identification and social Network construction through an online citizen science platform

Citizen science has grown rapidly in popularity in recent years due to its potential to educate and engage the public while providing a means to address a myriad of scientific questions. However, the rise in popularity of citizen science has also been accompanied by concerns about the quality of data emerging from citizen science research projects. We assessed data quality in the online citizen scientist platform Chimp&See, which hosts camera trap videos of chimpanzees (Pan troglodytes) and other species across Equatorial Africa. In particular, we compared detection and identification of individual chimpanzees by citizen scientists to that of experts with years of experience studying those chimpanzees. We found that citizen scientists typically detected the same number of individual chimpanzees as experts, but assigned far fewer identifications (IDs) to those individuals. Those IDs assigned, however, were nearly always in agreement with the IDs provided by experts. We applied the data sets of citizen scientists and experts by constructing social networks from each. We found that both social networks were relatively robust and shared a similar structure, as well as having positively correlated individual network positions. Our findings demonstrate that, although citizen scientists produced a smaller data set based on fewer confirmed IDs, the data strongly reflect expert classifications and can be used for meaningful assessments of group structure and dynamics. This approach expands opportunities for social research and conservation monitoring in great apes and many other individually identifiable species.

Football is becoming more predictable: Network analysis of 88 thousand matches in 11 major leagues

In recent years excessive monetization of football and professionalism among the players has been argued to have affected the quality of the match in different ways. On the one hand, playing football has become a high-income profession and the players are highly motivated; on the other hand, stronger teams have higher incomes and therefore afford better players leading to an even stronger appearance in tournaments that can make the game more imbalanced and hence predictable. To quantify and document this observation, in this work we take a minimalist network science approach to measure the predictability of football over 26 years in major European leagues. We show that over time, the games in major leagues have indeed become more predictable. We provide further support for this observation by showing that inequality between teams has increased and the home-field advantage has been vanishing ubiquitously. We do not include any direct analysis on the effects of monetization on football's predictability or therefore, lack of excitement, however, we propose several hypotheses which could be tested in future analyses.

opencc-by-4.0Oct 2021View details →

Logic Shrinkage: Learned FPGA Netlist Sparsity for Efficient Neural Network Inference [Artefact Evaluation]

Source code of paper "Logic Shrinkage: Learned FPGA Netlist Sparsity for Efficient Neural Network Inference" submitted to FPGA'22 for artefact evaluation.

Recognition of dominant driving factors behind sap flow of Liquidambar formosana based on back-propagation neutral network method

Aims: This study focused on the applicability of back-propagation (BP) neural networks in simulating sap flow (SF) using meteorological factors and a phenological index (PI) for Liquidambar formosana, a deciduous broad-leaf tree species in subtropical China, and thus providing a useful and promising alternative to traditional methods for transpiration prediction. Methods: Three-layered BP models with an architecture 4-10-1 (four neurons in the input layer, ten neurons in the hidden layer and one neuron in the output layer) were trained and tested using the Levenberg-Marquardt (LM) algorithm based on in situ observations of SF and concurrent microclimate at the Qianyanzhou Ecological Station, Jiangxi Province, Southeast China. The model performance was verified with testing data not used in model development. Results: The BP models with eight input combinations proved a satisfactory fit: the determination coefficients (R2) and fitting accuracies (Acc) (about 0.8 and 70%) were significantly higher than those of the multivariate linear regression (MLR) (about 0.5 and 50%), indicating their advantage in solving complex nonlinear problems involved in transpiration. In addition, the BP models showed a bit better performance by adding PI to the input family. The best BP model was achieved taking air temperature (Ta), relative humidity (RH), average net radiation (ANR) and PI as the input and sap flux density (vs) as the output, with maximum R2 and Acc as high as 0.95 and 90%, respectively. Conclusions: The BP models with input combination of Ta, RH, ANR and PI mirrored very well measured daily variations in vs. The results could be used to fine-tune sap flow estimation by Liquidambar formosana, and thus shed light on the eco-hydrological process related to transpiration for deciduous broad-leaf trees.

opennotspecifiedNov 2021View details →

FIGURE. TCS network inferred from ITS1 in Morphological and phylogenetic relations of members of the genus Coelastrella (Scenedesmaceae, Chlorophyta) from the Ural and Khentii Mountains (Russia, Mongolia)

FIGURE. TCS network inferred from ITS1 Coelastrella sequences. The area of a circle is proportional to the number of Coelastrella sequences available in GenBank database. Mutational events between haplotypes are indicated by hatch marks at branches. The network was inferred using the algorithm described by Clement et al. (2002).

opencc-by-4.0Jul 2018View details →

Screencast of the Lokahi2 Prototype: Search Engine with Interactive Knowledge Network Browser Extracted from Text

This video shows a recording of the prototype system Lokahi2. It supports concept surfing for interacting with the search engine, automated document tagging, exploring tags, generating tags for text, and changing depth and dimension of the graph.

opencc-by-4.0Nov 2021View details →

NicheNet networks update data

cf separate doc for explanation

opencc-zeroDec 2021View details →

Ecological network complexity scales with area

Larger geographical areas contain more species–an observation raised to a law in ecology. Less explored is whether biodiversity changes are accompanied by a modification of interaction networks. We use data from 32 spatial interaction networks from different ecosystems to analyse how network structure changes with area. We find that basic community structure descriptors (number of species, links, and links per species) increase with area following a power-law. Yet, the distribution of links per species varies little with area, indicating that the fundamental organization of interactions within networks is conserved. Our null model analyses suggest that the spatial scaling of network structure is determined by factors beyond species richness and the number of links. We demonstrate that biodiversity–area relationships can be extended from species counts to higher levels of network complexity. Therefore, the consequences of anthropogenic habitat destruction may extend from species loss to wider simplification of natural communities.

opencc-by-4.0Nov 2008View details →

Accuracy Study and Improvement of Network Simulation in the SimGrid Framework (datasets)

Data sets and code for https://hal.inria.fr/inria-00361031 and the Rescience replication. The content.org file is the content of the Rescience article from which all the commands for the replication can be found.

opencc-by-4.0Dec 2021View details →

AHI-CALIOP Collocated Data for Training and Validation of Cloud Masking Neural Networks

Collocated data between AHI at 2km resolution (nadir) and CALIOP 1km cloud product v4.20 used for training and validating cloud identification neural networks. The main training and validation data from 2019 is stored in monthly directories, whilst the collocated dataset used to compare the NN, JMA and BoM cloud mask performances is the file "superdf.h5". All collocated data is stored as .h5 files and was built using the Python Pandas package. In this archive, the data has been stored as compressed directories for each month or as a single compressed file in the case of "superdf.h5" using tar with bzip2 compression or just bzip2 compression respectively.

opencc-by-4.0Dec 2021View details →

Demodulation Combining 1D-CNN and Bi-LSTM Network over Strong Solar Wind Turbulence Channel

The data in this dataset is derived MATLAB simulation dataset by us to illustrate the GMSK signal demodualtion over strong solar wind turbulence using deep learing. Disclaimer The data provided in the files is provided as is. Despite our best efforts at filtering out potential issues, some information could be erroneous. Description of the dataset One file per is provided as a csv file with the following features: train_dataset:  Including data_kb2 and data_awgn. data_kb2 is the GMSK modulated fading signal data and lable data under the influence of strong solar wind turbulence, data_awgn is the GMSK modulated signal data and lable data under the influence of Gaussian white noise only.In which per mod_data and lable_data have 21000 csv file, respectively. And per csv file has 800 data. Among them ,the lable is the original binary number. The train_dataset is used to train neural network demodulator model. test_dataset:Including test_data and test_lable.The test_data and test_lable have 6 file respectively. And they are used to test the trained neural network demodulator model. test_data: kb2_Tb0d5_m1d4: GMSK data at the BTb=0.5  , solar wind turbulence scintillation index m=1.4. kb2_Tb0d5_m1d2:GMSK data at the BTb=0.5  , solar wind turbulence scintillation index m=1.2. kb2_Tb0d3_m1d4:GMSK data at  the BTb=0.3  , solar wind turbulence scintillation index m=1.4. kb2_Tb0d3_m1d4: GMSK data at the BTb=0.3  , solar wind turbulence scintillation index m=1.2. Awgn_Tb0d5:GMSK at data the BTb=0.5 under the influence of Gaussian white noise only. Awgn_Tb0d5:GMSK data at the BTb=0.3 under the influence of Gaussian white noise only. test_lable: kb2_Tb0d5_m1d4: GMSK lable data at the BTb=0.5 , solar wind turbulence scintillation index m=1.4. kb2_Tb0d5_m1d2:GMSK lable data at the BTb=0.5  , solar wind turbulence scintillation index m=1.2. kb2_Tb0d3_m1d4:GMSK labe data at  the BTb=0.3  , solar wind turbulence scintillation index m=1.4. kb2_Tb0d3_m1d4: GMSK labe data at the BTb=0.3  , solar wind turbulence scintillation index m=1.2. awgn_Tb0d5:GMSK lable data at the BTb=0.5 under the influence of Gaussian white noise only. awgn_Tb0d5:GMSK labe data at the BTb=0.3 under the influence of Gaussian white noise only.

opencc-by-4.0Dec 2021View details →

Data sets used in "Neural network emulation of the formation of organic aerosols based on the explicit GECKO-A chemistry model"

The training, validation, and testing data sets for toluene, dodecane, and alpha-pinene models described in the manuscript. A link to the manuscript will be added here when it becomes available. All trajectories in the data sets were generated using GECKO-A. The source code for using the data sets can be found at https://github.com/NCAR/gecko-ml

opencc-by-4.0Aug 2020View details →

Supporting Information: Environmental benefits of large-scale second-generation bioethanol production in the EU: An integrated supply chain network optimization and Life Cycle Assessment approache

This supporting information provides all input data of the model, the assumptions made and the literature and database references for the publication Environmental benefits of large-scale second-generation bioethanol production in the EU: An integrated supply chain network optimization and Life Cycle Assessment approache. The environmental data is based on life cycle assessments, with full information on the life cycle inventory and the results of the life cycle impact assessment based on the ReCiPe method. It also includes detailed results for all objective functions in all scenarios (optimization of 18 midpoints, 3 endpoints, and economic optimization in 5 tax scenarios and 2 feedstock scenarios), and detailed results of the sensitivity analysis and Pareto optimization.