Characteristics of PDF estimates for the bimodal distribution described by a binary mixture of two Gaussian distributions.

This dataset includes the data in the figures for the article entitled "The effect of a bimodal gravel size distribution on clean-out depth of sand", which is to be published in the Journal of Hydraulic Engineering. The authors are: Kuhnle, R. A., Smith, J. E., Wren, D. G., and Langendoen, E. L.The clogging of the gravel surface layer in gravel-bedded channels with finer sediment (< 2 mm) is a common problem that negatively impacts stream bed habitat quality. Most existing measurements of surface layer sediment clean-out depth are from experiments with narrow gravel size distributions; therefore, a new set of experiments was conducted to explore the effect of widely distributed bimodal gravel fractions on the clean-out depth of interstitial sand. Experiments were used to measure the terminal clean-out depth of 0.136 mm sand from bed material substrates composed of unimodal median diameter gravel of 7.16 mm and bimodal bed material with modes of 7.31 and 34.66 mm and a median diameter of 33.46 mm. It was found that the clean-out depth of sand from immobile gravel beds could be predicted by scaling the bed shear stress using the cumulative probability distribution of the elevations of the surface of the gravel bed.

Heermann et al_PLOSONEIn this file, data on food resources (zooplankton and macroinvertebrates) of perch can be found. Also given is total length of fish used in the analyses as well as the percentage composition of diet items and where available information about sex and maturation state of perch. Further, data derived from Procrustes superimposition describing shape of perch is shown.

Bimodal dataset on Inner Speech

Code available: https://github.com/LTU-Machine-Learning/Inner_Speech_EEG_FMRI

Publication available: https://www.biorxiv.org/content/10.1101/2022.05.24.492109v3

Abstract: The recognition of inner speech, which could give a `voice' to patients that have no ability to speak or move, is a challenge for brain-computer interfaces (BCIs). A shortcoming of the available datasets is that they do not combine modalities to increase the performance of inner speech recognition. Multimodal datasets of brain data enable the fusion of neuroimaging modalities with complimentary properties, such as the high spatial resolution of functional magnetic resonance imaging (fMRI) and the temporal resolution of electroencephalography (EEG), and therefore are promising for decoding inner speech. This paper presents the first publicly available bimodal dataset containing EEG and fMRI data acquired nonsimultaneously during inner-speech production. Data were obtained from four healthy, right-handed participants during an inner-speech task with words in either a social or numerical category. Each of the 8-word stimuli were assessed with 40 trials, resulting in 320 trials in each modality for each participant.
The aim of this work is to provide a publicly available bimodal dataset on inner speech, contributing towards speech prostheses.

Short Dataset description: The dataset consists of 1280 trials in each modality (EEG, FMRI). The stimuli contain 8 words, selected from 2 different categories (social, numeric): Social: child, daughter, father, wife Numeric: four, three, ten, six

There are 4 subjects in total: sub-01, sub-02, sub-03, sub-05. Initially, there were 5 participants, however, sub-04 data was rejected due to high fluctuations. Details of valid data are available in the file participants.tsv.

For questions please contact: foteini.liwicki@ltu.se

Dataset

This dataset was created by Mohamed Abubakr

Contents

The MERGE dataset is a collection of audio, lyrics, and bimodal datasets for conducting research on Music Emotion Recognition. A complete version is provided for each modality. The audio datasets provide 30-second excerpts for each sample, while full lyrics are provided in the relevant datasets. The amount of available samples in each dataset is the following:

• MERGE Audio Complete: 3554
• MERGE Audio Balanced: 3232
• MERGE Lyrics Complete: 2568
• MERGE Lyrics Balanced: 2400
• MERGE Bimodal Complete: 2216
• MERGE Bimodal Balanced: 2000

Additional Contents

Each dataset contains the following additional files:

• av_values: File containing the arousal and valence values for each sample sorted by their identifier;
• tvt_dataframes: Train, validate, and test splits for each dataset. Both a 70-15-15 and a 40-30-30 split are provided.

Metadata

A metadata spreadsheet is provided for each dataset with the following information for each sample, if available:

• Song (Audio and Lyrics datasets) - Song identifiers. Identifiers starting with MT were extracted from the AllMusic platform, while those starting with A or L were collected from private collections;
• Quadrant - Label corresponding to one of the four quadrants from Russell's Circumplex Model;
• AllMusic Id - For samples starting with A or L, the matching AllMusic identifier is also provided. This was used to complement the available information for the samples originally obtained from the platform;
• Artist - First performing artist or band;
• Title - Song title;
• Relevance - AllMusic metric representing the relevance of the song in relation to the query used;
• Duration - Song length in seconds;
• Moods - User-generated mood tags extracted from the AllMusic platform and available in Warriner's affective dictionary;
• MoodsAll - User-generated mood tags extracted from the AllMusic platform;
• Genres - User-generated genre tags extracted from the AllMusic platform;
• Themes - User-generated theme tags extracted from the AllMusic platform;
• Styles - User-generated style tags extracted from the AllMusic platform;
• AppearancesTrackIDs - All AllMusic identifiers related with a sample;
• Sample - Availability of the sample in the AllMusic platform;
• SampleURL - URL to the 30-second excerpt in AllMusic;
• ActualYear - Year of song release

Acknowledgements

This work is funded by FCT - Foundation for Science and Technology, I.P., within the scope of the projects: MERGE - DOI: 10.54499/PTDC/CCI-COM/3171/2021 financed with national funds (PIDDAC) via the Portuguese State Budget; and project CISUC - UID/CEC/00326/2020 with funds from the European Social Fund, through the Regional Operational Program Centro 2020.

Renato Panda was supported by Ci2 - FCT UIDP/05567/2020.

Dataset

This dataset was created by reham abobakr

Contents

In educational measurement, exploring the method of generating multiple high-quality parallel tests has become a research hotspot. One purpose of this research is to construct parallel forms item by item according to a seed test, using two proposed item selection heuristic methods [minimum parameters–information–distance method (MPID) and minimum information–parameters–distance method (MIPD)]. Moreover, previous research addressing test assembly issues has been limited mainly to situations in which the information curve of the item pool or seed test has a normal or skewed distribution. However, in practice, the distributions of information curves for tests are diverse. These include multimodal distributions, the most common type of which is the bimodal distribution. Therefore, another main aim of this article is to extend the information curves of unimodal distributions to bimodal distributions. Thus, this study adopts simulation research to compare the results of two item, response, theory (IRT)-based item matching methods (MPID and MIPD) using different information curve distributions for item pools or seed tests. The results show that the MPID and MIPD methods yield rather good performance in terms of both two statistical targets when the information curve has a unimodal distribution, and two new methods yield better performance than two existing methods in terms of test information functions target when the information curve has a bimodal distribution.

Dataset

This dataset was created by Mohamed Abubakr

Contents

Data used to create sinusoidal functions that approximate month-to-month rainfall patterns on a global scale and are used to find whether locations experience one or two rainfall seasons per year.

Small-angle X-ray scattering curves of bimodal SiO2 nanoparticles and the required Python algorithms for model fitting. The results of the fits (size distribution curves, parameters etc.) are also included.

Dataset related to a bimodal fluorescent reporter for the simultaneous and highly specific detection of two independent protease families of biological importance, based on a dual, multiplexing, peptide FRET system.

CROP GARDEN RAWThe raw data of 2012 experiment for yield canalization.Data collected at the field and in the lab as well.2004 rawData was collected by Yaniv Semel at 2004, this data was reanalyzed for this paper.

GPU-Disaggregated DLRM Trace Dataset

This dataset records latency-sensitive inference instances for GPU-disaggregated serving of deep learning recommendation models. It contains per-instance resource reservations and life cycle timestamps for scheduling analysis and capacity planning.

This dataset represents a groundbreaking trace collection from production GPU-disaggregated serving systems for Deep Learning Recommendation Models (DLRMs), accompanying the NSDI'25 paper on GPU-disaggregated serving at scale. The dataset captures real-world operational characteristics of inference services in a large-scale production environment, providing invaluable insights into resource allocation patterns, temporal dynamics, and system behavior for latency-sensitive ML workloads.

Scope

• Total rows: 23871.
• Unique apps: 156.
• Role split, {'CN': 16485, 'HN': 7386}.

Key fields

Instance ID. Role. Application group. Requests and limits for CPU, GPU, RDMA, memory, and disk. Density cap per node. Creation, scheduling, and deletion timestamps relative to the trace start.

High-level observations

• All workloads are marked latency sensitive. Instances are typically long-running with high priority, as stated by the authors.
• Scheduling delay distribution and runtime distribution are included in the figures. Concurrency over time gives a view of system load.
• RDMA percentage and max instances per node expose placement constraints that influence packing on heterogeneous nodes.

🎯 Key Characteristics

Scale and Scope

• Total Instances: 23,871 inference instances
• Services: 156 unique inference services (applications)
• Workload Type: 100% Latency-Sensitive (LS) workloads
• Priority Level: High-priority, long-running inference instances
• System Architecture: GPU-disaggregated architecture separating compute and GPU resources

Instance Distribution

• CPU Nodes (CN): 16,485 instances (69.1%)
  • Pure CPU-based inference workloads
  • No GPU allocation
  • Lower RDMA requirements (mean: 3.4%)
• Heterogeneous GPU Nodes (HN): 7,386 instances (30.9%)
  • GPU-accelerated inference workloads
  • All instances allocated exactly 1 GPU
  • Higher RDMA requirements (mean: 20.5%)

🔍 Key Insights

Workload Heterogeneity

• Clear bimodal distribution between CPU and GPU workloads
• CN instances optimized for CPU-intensive operations
• HN instances balanced for GPU acceleration with supporting CPU resources

Resource Efficiency

• Tight coupling between CPU and memory allocation (correlation: 0.97)
• Independent scaling of GPU resources from CPU/memory
• RDMA bandwidth scaled based on disaggregation communication needs

Production Patterns

• All workloads classified as latency-sensitive
• High-priority, long-running inference services
• Immediate scheduling indicates sufficient resource availability

Disaggregation Benefits

• Efficient resource utilization through separation of concerns
• CN nodes handle CPU-intensive preprocessing/postprocessing
• HN nodes focus on GPU-accelerated model inference
• RDMA enables efficient data movement between disaggregated components

📈 Research Applications

This dataset enables research in:

• Resource Allocation: Optimal scheduling strategies for disaggregated systems
• Performance Modeling: Understanding latency-throughput tradeoffs
• System Design: Architectural decisions for ML serving infrastructure
• Workload Characterization: Production DLRM inference patterns
• Capacity Planning: Resource provisioning for ML workloads
• Fault Tolerance: Instance distribution and anti-affinity strategies

🎓 Academic Contribution

This dataset represents one of the first publicly available production traces for GPU-disaggregated DLRM serving, providing:

• Real-world validation data for system research
• Baseline for performance comparisons
• Foundation for reproducible research in ML systems
• Insights into production-scale ML infrastructure

This dataset provides a unique window into production GPU-disaggregated systems, offering researchers and practitioners valuable insights for advancing the field of large-scale ML serving infrastructure.

This is the full data set of the PER associative conditioned experiments to test the effect of Synchronicity level, temporal order and intensity on the bimodal learning of honey bees.

Dataset Card for "bimodal-lmpa-shuffled"

More Information needed

Fundamental to biological decision-making is the ability to generate bimodal expression patterns where 2 alternate expression states simultaneously exist. Here, we use a combination of single-cell analysis and mathematical modeling to examine the sources of bimodality in the transcriptional program controlling HIV’s fate decision between active replication and viral latency. We find that the HIV transactivator of transcription (Tat) protein manipulates the intrinsic toggling of HIV’s promoter, the long terminal repeat (LTR), to generate bimodal ON-OFF expression and that transcriptional positive feedback from Tat shifts and expands the regime of LTR bimodality. This result holds for both minimal synthetic viral circuits and full-length virus. Strikingly, computational analysis indicates that the Tat circuit’s noncooperative “nonlatching” feedback architecture is optimized to slow the promoter’s toggling and generate bimodality by stochastic extinction of Tat. In contrast to the standard Poisson model, theory and experiment show that nonlatching positive feedback substantially dampens the inverse noise-mean relationship to maintain stochastic bimodality despite increasing mean expression levels. Given the rapid evolution of HIV, the presence of a circuit optimized to robustly generate bimodal expression appears consistent with the hypothesis that HIV’s decision between active replication and latency provides a viral fitness advantage. More broadly, the results suggest that positive-feedback circuits may have evolved not only for signal amplification but also for robustly generating bimodality by decoupling expression fluctuations (noise) from mean expression levels.

The amplitude values at the 75% and 84% thresholds (in dB SPL for A and Av; Cm for V and Va) in the unimodal and bimodal experiments (columns) for all animals (rows 1–4).Comparison of threshold values for uni- and bimodal experiments.

This table contains a subset of the results of a correlation of the ROSAT All-Sky Survey (RASS) with the April 1997 release of the VLA 20-cm Faint Images of the Radio Sky at Twenty cm (FIRST: CDS Cat. ) Catalog. It focusses on the analysis of the 843 X-ray sources which have unique radio counterparts. The majority of these objects (84%) have optical counterparts on the POSS 1 plates. Approximately 30% have been previously classified and the authors obtain new spectroscopic classifications for 85 sources by comparison with the ongoing FIRST Bright Quasar Survey and 106 additional sources from their own new spectroscopic data. Approximately 51% of the sources are presently classified, and the majority of the unclassified objects are optically faint. The newly classified sources are generally radio weak, exhibiting properties intermediate with previous samples of radio- and X-ray-selected AGN. This also holds for the subsample of 71 BL Lacs which includes many intermediate objects. The 146 quasars show no evidence for a bimodal distribution in their radio-loudness parameter, indicating that the supposed division between radio-quiet and radio-loud AGN may not be real. The X-ray and radio luminosities are correlated over two decades in radio luminosity, spanning the radio-loud and radio-quiet regimes, with radio-quiet quasars showing a linear correlation between the two luminosities. Many of the sources show peculiar or unusual properties which call for more detailed follow-up observations. In their paper (Table 2), the authors also give the X-ray and radio data for the 518 X-ray sources for which more than one radio object is found. Because of the difficulties inherent in identifying optical counterparts to these complex sources, they do not consider these data in the current analysis, and they are not included in the present table (but are available at http://cdsarc.u-strasbg.fr/ftp/cats/J/A+A/356/445/). This table was created by the HEASARC in March 2012 based on CDS Catalog J/A+A/356/445 file table1.dat, the list of ROSAT All-Sky Survey sources with single FIRST matches. This is a service provided by NASA HEASARC .

Background and Objective: Respiratory diseases are among the most significant causes of morbidity and mortality worldwide, causing substantial strain on society and health systems. Over the last few decades, there has been increasing interest in the automatic analysis of respiratory sounds and electrical impedance tomography (EIT). Nevertheless, no publicly available databases with both respiratory sound and EIT data are available.

Methods: In this work, we have assembled the first open-access bimodal database focusing on the differential diagnosis of respiratory diseases (BRACETS: Bimodal Repository of Auscultation Coupled with Electrical Impedance Thoracic Signals). It includes simultaneous recordings of single and multi-channel respiratory sounds and EIT. Furthermore, we have proposed several machine learning-based baseline systems for automatically classifying respiratory diseases in six distinct evaluation tasks using respiratory sound and EIT (A1, A2, A3, B1, B2, B3). These tasks included classifying respiratory diseases at sample and subject levels. The performance of the classification models was evaluated using a 5-fold cross-validation scheme (with subject isolation between folds).

Results: The resulting database consists of 1097 respiratory sounds and 795 EIT recordings acquired from 78 adult subjects in two countries (Portugal and Greece). In the task of automatically classifying respiratory diseases, the baseline classification models have achieved the following average balanced accuracy: Task A1 - 77.9±13.1%; Task A2 - 51.6±9.7%; Task A3 - 38.6±13.1%; Task B1 - 90.0±22.4%; Task B2 - 61.4±11.8%; Task B3 - 50.8±10.6%.

Conclusion: The creation of this database and its public release will aid the research community in developing automated methodologies to assess and monitor respiratory function, and it might serve as a benchmark in the field of digital medicine for managing respiratory diseases. Moreover, it could pave the way for creating multi-modal robust approaches for that same purpose.