Historical Dataset Research Articles

Density data for Lake Erie benthic invertebrate assemblages from 1930 to 2019.

Benthic invertebrates are important trophic links in food webs and useful bioindicators of environmental conditions, but long-term benthic organism abundance data across broad geographic areas are rare and historic datasets are often not readily accessible. This dataset provides densities of benthic macroinvertebrates collected from 1930 to 2019 during surveys in Lake Erie, a Laurentian Great Lake. The surveys were funded by the governments of the United States and Canada to investigate the status and changes in the benthic community. From the total of 21 lake-wide and basin-wide benthic surveys conducted in Lake Erie from 1929 to 2019, we were able to acquire data for 17 surveys, including species-level data for 10 surveys and data by higher taxonomic groups for seven surveys. Our amassed Lake Erie dataset includes data from 11 surveys (including five with species-level data) conducted in the western basin in 1930-2019, seven surveys (six with species-level data) in the central basin, and eight surveys (seven with species-level data) in the eastern basin (1973-2019). This Lake Erie dataset represents the most extensive temporal dataset of benthic invertebrates available for any of the Laurentian Great Lakes. Benthic samples were collected using Ponar or Shipek bottom dredges and taxa densities were calculated as individuals per square meter using the area of the dredge. Density data are provided for taxa in the Annelida, Arthropoda, Mollusca, Cnidaria, Nemertea, and Platyhelminthes phyla. Current taxonomy was used for most groups but, in a few cases, older taxonomic names were used for consistency with historical data. Analysis of this dataset indicates that eutrophication, water quality improvement, and dreissenid introduction were the major drivers of changes in the benthic community in the western basin, while hypoxia was a major factor in the central basin, and dreissenid introduction was the most important driver in the eastern basin. Considering the rarity of high taxonomic resolution long-term benthic data for lake ecosystems, this dataset could be useful to explore broader aspects of ecological theory, including effects of eutrophication, hypoxia, invasive species, and other factors on community organization, phylogenetic and functional diversity, and spatial and temporal scales of variation in community structure. In addition, the dataset could be useful for studies on individual species, including abundance and distribution, species co-occurrence, and how the patterns of dominance and rarity change over space and time. Use of this dataset for academic or educational purposes is encouraged as long as this data paper is properly cited.

Evaluation of Climatological Precipitation Datasets and Their Hydrological Application in the Hablehroud Watershed, Iran

Hydrological modeling is essential for runoff simulations in line with climate studies, especially in remote areas with data scarcity. Advancements in climatic precipitation datasets have improved the accuracy of hydrological modeling. This research aims to evaluate the APHRODITE, PERSIANN-CDR, and ERA5-Land climatic precipitation datasets for the Hablehroud watershed in Iran. The datasets were compared with interpolated ground station precipitation data using the inverse distance weighted (IDW) method. The variable infiltration capacity (VIC) model was utilized to simulate runoff from 1992 to 1996. The results revealed that the APHRODITE and PERSIANN-CDR datasets demonstrated the highest and lowest accuracy, respectively. The sensitivity of the model was analyzed using each precipitation dataset, and model calibration was performed using the Kling–Gupta efficiency (KGE). The evaluation of daily runoff simulation based on observed precipitation indicated a KGE value of 0.78 and 0.76 during the calibration and validation periods, respectively. The KGE values at the daily time scale were 0.64 and 0.77 for PERSIANN-CDR data, 0.62 and 0.75 for APHRODITE precipitation data, 0.50 and 0.66 for ERA5-Land precipitation data during the calibration and validation periods, respectively. These results indicate that despite varying sensitivity, climatic precipitation datasets present satisfactory performance, particularly in poorly gauged basins with infrequent historical datasets.

Power Demand Forecasting Using ANN and Prophet Models for the Load Despatch Center in Andhra Pradesh, India

This paper uses various data variables to develop and analyze ANN and Prophet models for power demand forecasting in Andhra Pradesh, India. The electricity power consumption in Andhra Pradesh was about 51,756.000 GWh in 2021. Currently, there is a great emphasis on saving power. Power Demand Forecasting is creating much interest, and many models, such as artificial neural networks combined with other techniques based on real-life phenomena, are used and tested. These models have become an essential part of the power and energy sector. This paper considered specific time-series analysis methods and deep-learning techniques for short-term power demand forecasting. This paper also analyzes and compares results between the prophet and ANN models to predict power demand in Andhra Pradesh, India. Our results comparatively revealed the model's appropriateness for the problem. Both models performed well in three performance metrics: accuracy, generalization, and robustness. However, the AI model exhibits better accuracy than Prophet for the historical data set. The time taken for model fitting is also comparatively less for the AI models. The forecast accuracy of the electricity was in the range of 95 to 97.65.

Algoritma Gated Recurrent Unit untuk Prediksi Harga Indeks Penutupan Saham LQ45

The Indonesia Stock Exchange (IDX) states that stocks, including LQ45 stocks, which constitute the stock market index for the IDX, have become one of the preferred investment options for the public. Investors need to have accurate analysis and information to gain significant profits as stock prices fluctuate due to company performance, industry factors, changes in interest rates, liquidity, global market conditions, market sentiment, and investor psychology. The Gated Recurrent Unit algorithm is suitable for application on historical stock data sets because they are time series data, can be computed and compared on a numerical scale. This algorithm is a variant of the Long Short-Term Memory algorithm or other types of processing modules for Recurrent Neural Networks. The data set used consists of closing price data or close features, comprising a training data set of 4,406 data and a test data set of 1,889 data that have undergone data preparation using various techniques, including data cleansing, data scrubbing, data splitting, data normalization, and data reshaping. The results showed that the Gated Recurrent Unit algorithm is the right strategy because it obtains a good evaluation of model performance, namely MSE of 0.0009; RMSE of 0.17325 and MAE of 0.0207.

TIME SERIES FORECASTING MODEL FOR SOLVING COLD START PROBLEM VIA TEMPORAL FUSION TRANSFORMER

Time series forecasting is an important tool in many businesses. It can range from efficiently allocating resources for web traffic, predicting patient needs for staffing requirements, to forecasting a company's product sales. A particular use case, known as "cold start" forecasting, involves making predictions for time series that have little or no historical data, like a new product just entering the retail market. The key assumption of cold start forecasting is that products with similar characteristics should have similar time series trajectories. In such scenarios, traditional forecasting models that heavily rely on past observations may face challenges, necessitating the development of innovative approaches that can effectively make predictions in the absence of a substantial historical dataset. In this paper, Temporal fusion transformer neural network architecture was applied for solving cold start time series forecasting task. Modeling of the method was based on the use of a dataset contained in an open repository. After the preprocessing procedures, the dataset has about 370 time series, each of which has different length of series and has one categorical feature. Categorical feature have only 4 types of different values. For model training was performed to search for optimal hyperparameters across such parameters as: number of attention heads, learning rate, dropout percentage and hidden size.Model performed pretty well on this task. For model comparison were chosen metrics: MAE, RMSE, SMAPE. As can be seen from comparison with such popular models as DeepAR and LSTM, the proposed approach demonstrated the smallest forecasting error. Only one downside is that it can have more problems with anomalies in time series than DeepAR. But at the same time still provide interpretability of results.

A CNN-GRU Hybrid Model for Predicting Airport Departure Taxiing Time

Scene taxiing time is an important indicator for assessing the operational efficiency of airports as well as green airports, and it is also a fundamental parameter in flight regularity statistics. The accurate prediction of taxiing time can help decision makers to further optimize flight pushback sequences and improve airport operational efficiency while increasing flight punctuality. In this paper, we propose a hybrid deep learning model for departure taxiing time prediction based on the new influence factors of taxiing time. Taking Pudong International Airport as the research object, after analyzing the scene operation mode, we construct the origin–destination pairs (ODPs) with stand groups and runways and then propose two structure-related factors, corridor departure flow and departure flow proportion of ODP, as the new features. Based on the new feature set, we construct a departure taxiing dataset for training the prediction model. Then, a departure taxiing time prediction model based on convolutional neural networks (CNNs) and gated recurrent units (GRUs) is proposed, which uses a CNN model to extract the high-dimensional features from the taxiing data and then inputs them to a GRU model for taxiing time prediction. Finally, we conduct a series of comparison experiments on the historical taxiing dataset of Pudong Airport. The prediction results show that the proposed hybrid prediction model has the best performances compared with other deep learning models, and the proposed structure-related features have high correlations with departure taxiing time. The prediction results of taxiing time for different ODPs also verify the generalizability of the proposed model.

How comparable are patient outcomes in the “real-world” with populations studied in pivotal AML trials?

Despite an increasing desire to use historical cohorts as “synthetic” controls for new drug evaluation, limited data exist regarding the comparability of real-world outcomes to those in clinical trials. Governmental cancer data often lacks details on treatment, response, and molecular characterization of disease sub-groups. The Australasian Leukaemia and Lymphoma Group National Blood Cancer Registry (ALLG NBCR) includes source information on morphology, cytogenetics, flow cytometry, and molecular features linked to treatment received (including transplantation), response to treatment, relapse, and survival outcome. Using data from 942 AML patients enrolled between 2012–2018, we assessed age and disease-matched control and interventional populations from published randomized trials that led to the registration of midostaurin, gemtuzumab ozogamicin, CPX-351, oral azacitidine, and venetoclax. Our analyses highlight important differences in real-world outcomes compared to clinical trial populations, including variations in anthracycline type, cytarabine intensity and scheduling during consolidation, and the frequency of allogeneic hematopoietic cell transplantation in first remission. Although real-world outcomes were comparable to some published studies, notable differences were apparent in others. If historical datasets were used to assess the impact of novel therapies, this work underscores the need to assess diverse datasets to enable geographic differences in treatment outcomes to be accounted for.

A multi-task deep learning framework for forecasting sparse demand of demand responsive transit

Demand Responsive Transport (DRT) is a flexible transportation service that adjusts its routes and schedules in real time based on user requests. Real-time demand forecasting has become a timely issue since the resolution of transportation planning has increased to provide flexible services to users. Although many previous studies have been conducted to design the demand forecasting model using the mobility service historical data, there are more considerations when predicting real-time demand for on-demand mobility services. Therefore, this study aims to develop a demand forecasting model for DRT services addressing two issues. First, repeated calls for service and abandoned requests are converted into the unit of demand to capture served and potential demand. Second, given the sparsity of DRT historical data, a multi-task deep learning model is proposed to predict both the probability and volume of demand simultaneously to mitigate the data sparsity problem. A zero-inflated loss function was introduced to consider the skewed demand distribution. To attain our goals, the historical dataset of the I-MOD service trial in Yeongjongdo, Inchon, Korea was used. The performance of the proposed model is compared against several baseline models, demonstrating its superiority in forecasting demand for mobility services. The results show that a multi-task structure can alleviate the data sparsity problem by excluding where demand is not generated from the forecasting problem. The proposed models can provide insights into real-time demand forecasting with sparse demand.

An energy-efficient path planning method for unmanned surface vehicle in a time-variant maritime environment

This research presents a novel method for energy-efficient path planning, aiming to enhance the endurance of unmanned surface vehicle (USV). The proposed method combines the locking sweeping (LS) method, gradient descent method, coastline expansion method, and energy consumption functions. The efficiency of constructing the energy consumption potential map was improved by optimising the LS structure. Moreover, maintaining a user-configurable distance between the USV and the coastline ensures safe paths with minimal energy consumption. The performance of the proposed method was evaluated using multiple simulations involving high-resolution electronic nautical charts and a historical time-variant sea current dataset. The results demonstrate the practicality of the method, and effectively address path planning challenges in a time-variant maritime environment.

Critical care management of acute intoxications, dynamics and changes over time: a cohort study

Acute intoxications and poisonings are a relevant cause for ICU admission of critically ill patients. This study aimed to determine the characteristics of intoxicated patients in a tertiary center medical ICU in Austria over time and to investigate parameters associated with ICU mortality. This study was a retrospective data analysis including adult ICU patients from the years 2007 to 2021. In addition to ICU documentation, pre-hospital, and emergency department documents as well as autopsy reports were utilized. In an exploratory subanalysis, we compared these findings to a historical dataset from our facility from 1992 to 1996. We identified 581 cases admitted to the medical ICU because of acute poisoning (2007–2021), of which 45% were female and 46.6% were mixed intoxications. Suicidal intent was the primary cause of intoxication (48.2%) and ICU length of stay was median 1.2 days. The majority of deceased patients received pre-hospital mechanical CPR. Primary and secondary poison/toxin removal modalities were used in 29.9% and 11.7% of cases, whereas antidotes were administered in 54.4%. Comparing the data with a historical cohort (n = 168), we found a shift in primary detoxification away from gastric lavage and an increase in alternative secondary poison/toxin removal techniques. The ICU mortality was 4.1% and 4.2% in the present and historic cohort, respectively. Pre-existing psychiatric illnesses increased from 49% in the historic to 69% in the present cohort. Psychiatric illness predisposes patients to severe intoxications necessitating ICU care, thus increasing prevention measures seems warranted. Females did present with a different spectrum of intoxications compared to males. ICU mortality remained low over time and most deceased patients had a grim prognosis already on ICU arrival.

Self-healing control of abnormal conditions for fused magnesium furnace based on data augmentation and improved JITL

Abnormal conditions that arise during the fused magnesia smelting process (FMSP) can negatively impact product quality and increase energy consumption. However, due to the process's nonlinear and time-varying nature and the limited availability of abnormal condition data, researching effective self-healing control measures has been severely constrained. Therefore, this paper proposes a fused magnesium furnace (FMF) abnormal condition self-healing control method based on data augmentation and improved just-in-time learning (JITL). Firstly, data augmentation technology is used to generate virtual abnormal condition samples to expand the historical dataset. Then, an improved JITL model is constructed to solve the self-healing measures corresponding to abnormal conditions. Experimental results demonstrate that the proposed method can effectively restore the FMF to normal conditions and outperforms existing methods.

Prevalence of resistance-associated viral variants to the HIV-specific broadly neutralising antibody 10-1074 in a UK bNAb-naïve population.

Broadly neutralising antibodies (bNAbs) targeting HIV show promise for both prevention of infection and treatment. Among these, 10-1074 has shown potential in neutralising a wide range of HIV strains. However, resistant viruses may limit the clinical efficacy of 10-1074. The prevalence of both de novo and emergent 10-1074 resistance will determine its use at a population level both to protect against HIV transmission and as an option for treatment. To help understand this further, we report the prevalence of pre-existing mutations associated with 10-1074 resistance in a bNAb-naive population of 157 individuals presenting to UK HIV centres with primary HIV infection, predominantly B clade, receiving antiretroviral treatment. Single genome analysis of HIV proviral envelope sequences showed that 29% of participants' viruses tested had at least one sequence with 10-1074 resistance-associated mutations. Mutations interfering with the glycan binding site at HIV Env position 332 accounted for 95% of all observed mutations. Subsequent analysis of a larger historic dataset of 2425 B-clade envelope sequences sampled from 1983 to 2019 revealed an increase of these mutations within the population over time. Clinical studies have shown that the presence of pre-existing bNAb mutations may predict diminished therapeutic effectiveness of 10-1074. Therefore, we emphasise the importance of screening for these mutations before initiating 10-1074 therapy, and to consider the implications of pre-existing resistance when designing prevention strategies.

Prediction of ground water quality in western regions of Tamilnadu using LSTM network

Assessing and safeguarding groundwater quality is critical for sustaining life in water-scarce regions like western Tamil Nadu. The motivation behind this study stems from the pressing need to address water quality challenges in a region grappling with scarcity. Despite existing efforts, a notable research gap exists in predictive tools that comprehensively capture the nuanced temporal variations and trends in groundwater quality. This is where the LSTM network steps in, showcasing exceptional accuracy in short-term predictions and discerning long-term trends. This research uses Long Short-Term Memory (LSTM) networks, a variant of recurrent neural networks, to predict groundwater quality in South Indian Regions, especially in Tamil Nadu. Extensive data, encompassing parameters such as pH, dissolved oxygen, turbidity, and various chemical constituents, were gathered over an extended timeframe. The LSTM model was then trained on this historical dataset, factoring in temporal dependencies and seasonality inherent in groundwater quality data. The validation process rigorously tests the LSTM model against actual groundwater quality measurements. The results were impressive, as the model demonstrated a remarkable ability to unravel the complex variations in groundwater quality.

Role of household energy efficiency in shaping policy directives toward clean electricity transition in the United States: A nonparametric multivariate QQR approach

Despite the growing interest in energy transition policies, electricity generation in the U.S. remains heavily dependent on natural gas and coal. In this paper, we investigate the role of household energy efficiency in shaping policy directives toward a clean electricity transition in the U.S. through the use of a novel nonparametric multivariate quantile on quantile regression (M−QQR) over the period 1970 to 2040. The empirical results reveal that household energy-related efficiency has a positive effect on clean electricity transition across quantiles. This implies that household energy efficiency promotes clean electricity transition in the U.S. The results also show that energy expenditure and intensity of energy-related CO2 emissions are negatively related to clean electricity transition. However, the role of energy expenditures in the higher quantiles is positive. These findings align with the sensitivity and robustness analyses. This study offers significant contributions: First, the recently extended bivariate quantile-on-quantile regression to a multivariate case is applied. Second, we use historical and forecast datasets that span over seven decades. Overall, this study suggests the need for policymakers to utilize energy efficiency measures to stimulate transition towards clean electricity.

Maximizing efficiency in sunflower breeding through historical data optimization

Genomic selection (GS) has become an increasingly popular tool in plant breeding programs, propelled by declining genotyping costs, an increase in computational power, and rediscovery of the best linear unbiased prediction methodology over the past two decades. This development has led to an accumulation of extensive historical datasets with genotypic and phenotypic information, triggering the question of how to best utilize these datasets. Here, we investigate whether all available data or a subset should be used to calibrate GS models for across-year predictions in a 7-year dataset of a commercial hybrid sunflower breeding program. We employed a multi-objective optimization approach to determine the ideal years to include in the training set (TRS). Next, for a given combination of TRS years, we further optimized the TRS size and its genetic composition. We developed the Min_GRM size optimization method which consistently found the optimal TRS size, reducing dimensionality by 20% with an approximately 1% loss in predictive ability. Additionally, the Tails_GEGVs algorithm displayed potential, outperforming the use of all data by using just 60% of it for grain yield, a high-complexity, low-heritability trait. Moreover, maximizing the genetic diversity of the TRS resulted in a consistent predictive ability across the entire range of genotypic values in the test set. Interestingly, the Tails_GEGVs algorithm, due to its ability to leverage heterogeneity, enhanced predictive performance for key hybrids with extreme genotypic values. Our study provides new insights into the optimal utilization of historical data in plant breeding programs, resulting in improved GS model predictive ability.

Machine learning motivated data imputation of storm data used in coastal hazard assessments

In the Coastal Hazards System's (CHS) Probabilistic Coastal Hazard Analysis (PCHA) framework developed by the United States Army Corps of Engineers (USACE), historical records of tropical cyclone parameters have been used as data sources for statistical analysis, including fitting marginal distributions and measuring correlations between storm parameters. One limitation of the available historical databases is that observations of central pressure and radius of maximum winds are not available for a large number of storms. This may adversely affect the results of statistical analyses used to develop hazard curves. This study uses machine learning techniques to develop a data imputation method to “fill in” missing storm parameter records in historical datasets used for Joint Probability Method (JPM)-based coastal hazard analysis such as the USACE's CHS-PCHA. Specifically, Gaussian process regression (GPR) and artificial neural network (ANN) models are investigated as candidate machine learning-derived data imputation models, and the performance of different model parameterizations is assessed. Candidate imputation models are compared against existing statistical relationships. The effect of the data imputation process on statistical analyses (marginal distributions and correlation measures) is also evaluated for a series of example coastal reference locations.

Interdependencies in Electrode Manufacturing: A Comprehensive Study Based on Design Augmentation and Explainable Machine Learning

AbstractElectrode manufacturing, as the core of battery cell production, is a complex process chain with a large number of interrelated parameters. An in‐depth understanding of the processes, their relevant parameters, and the resulting effects on intermediate and final product properties can accelerate the transition toward quality‐oriented, efficient battery cell production. Given the complexity of the process chain, data‐driven models have emerged as promising solutions for analyzing the existing interdependencies. The accuracy and effectiveness of these models significantly depend on the quality and comprehensiveness of the underlying data. With a low‐quality dataset, there is an increased risk of drawing inaccurate conclusions or generating misleading results. This article aimed to demonstrate a use case for the evaluation and enhancement of historical datasets to provide a statistically robust foundation for the development of machine learning models. The study was based on pilot‐scale anode manufacturing and covered variations in the coating, drying, and calendering processes. The key intermediate product and process parameters were used to predict two primary target variables: adhesion strength and discharge capacity at different C‐rates. To gain a better understanding of the analyzed interdependencies, explainable machine learning methods were adopted.

Enhancing Maritime Navigational Safety: Ship Trajectory Prediction Using ACoAtt–LSTM and AIS Data

Predicting ship trajectories plays a vital role in ensuring navigational safety, preventing collision incidents, and enhancing vessel management efficiency. The integration of advanced machine learning technology for precise trajectory prediction is emerging as a new trend in sophisticated geospatial applications. However, the complexity of the marine environment and data quality issues pose significant challenges to accurate ship trajectory forecasting. This study introduces an innovative trajectory prediction method, combining data encoding representation, attribute correlation attention module, and long short-term memory network. Initially, we process AIS data using data encoding conversion technology to improve representation efficiency and reduce complexity. This encoding not only preserves key information from the original data but also provides a more efficient input format for deep learning models. Subsequently, we incorporate the attribute correlation attention module, utilizing a multi-head attention mechanism to capture complex relationships between dynamic ship attributes, such as speed and direction, thereby enhancing the model’s understanding of implicit time series patterns in the data. Finally, leveraging the long short-term memory network’s capability for processing time series data, our approach effectively predicts future ship trajectories. In our experiments, we trained and tested our model using a historical AIS dataset. The results demonstrate that our model surpasses other classic intelligent models and advanced models with attention mechanisms in terms of trajectory prediction accuracy and stability.

Short-term association between air temperature and mortality in seven cities in Norway: A time series analysis.

The association between ambient air temperature and mortality has not been assessed in Norway. This study aimed to quantify for seven Norwegian cities (Oslo, Bergen, Stavanger, Drammen, Fredrikstad, Trondheim and Tromsø) the non-accidental, cardiovascular and respiratory diseases mortality burden due to non-optimal ambient temperatures. We used a historical daily dataset (1996-2018) to perform city-specific analyses with a distributed lag non-linear model with 14 days of lag, and pooled results in a multivariate meta-regression. We calculated attributable deaths for heat and cold, defined as days with temperatures above and below the city-specific optimum temperature. We further divided temperatures into moderate and extreme using cut-offs at the 1st and 99th percentiles. We observed that 5.3% (95% confidence interval (CI) 2.0-8.3) of the non-accidental related deaths, 11.8% (95% CI 6.4-16.4) of the cardiovascular and 5.9% (95% CI -4.0 to 14.3) of the respiratory were attributable to non-optimal temperatures. Notable variations were found between cities and subgroups stratified by sex and age. The mortality burden related to cold dominated in all three health outcomes (5.1%, 2.0-8.1, 11.4%, 6.0-15.4, and 5.1%, -5.5 to 13.8 respectively). Heat had a more pronounced effect on the burden of respiratory deaths (0.9%, 0.2-1.0). Extreme cold accounted for 0.2% of non-accidental deaths and 0.3% of cardiovascular and respiratory deaths, while extreme heat contributed to 0.2% of non-accidental and to 0.3% of respiratory deaths. Most of the burden could be attributed to the contribution of moderate cold. This evidence has significant implications for enhancing public-health policies to better address health consequences in the Norwegian setting.

A data decomposition and attention mechanism-based hybrid approach for electricity load forecasting

An accurate and reliable prediction of future energy patterns is of utmost significance for the smooth operation of several related activities such as capacity or generation unit planning, transmission network optimization, better resources availability, and many more. With the availability of historical load datasets through smart grid systems, artificial intelligence and machine learning-based techniques have been extensively developed for achieving the desired objectives. However, effectively capturing strong randomness and non-linear fluctuations in the load time-series remains a critical issue that demands concrete solutions. Considering this, the current research proposes a hybrid approach amalgamating data smoothing and decomposition strategy with deep neural models for improving forecasting results. Moreover, an attention mechanism is integrated to capture relevant portions of the time series, thus achieving the desired ability to capture long-term dependencies among load demand observations. This integration enhances the prediction and generalization capabilities of the proposed model. To validate the performance benefits achieved by the proposed approach, a comparative evaluation is conducted with state-of-the-art neural-based load series prediction models. The performance assessment is carried out on a novel real-world dataset of five southern states of India, and the superiority of the proposed in capturing load time-series variations is well observed and demonstrated in terms of several performance indicators.