Extreme Tree Research Articles

Hybrid Model of Machine Learning Refractory Data Prediction Based on IoT Smart Cities

With the advent of the digital age in recent years, the application of artificial intelligence in urban Internet of Things (IoT) systems has become increasingly important. The concept of smart cities has gradually formed, and smart firefighting under the smart city system has also become important. The method of machine learning is now applied in various fields, but seldom to the data prediction of smart firefighting. Various types of applications including data applications of machine learning algorithms in smart firefighting have yet to be explored. In this article, we propose using machine learning algorithms to predict building fire-resistance data, aiming to provide more theoretical and technical support for IoT smart cities. This article adopts the fire-resistance data of building beam components in a real fire environment, using three integrated machine learning algorithms, Extreme random Tree (ET), AdaBoost, and Gradient Boosting Machine (GBM), and the grey wolf optimization algorithm to optimize. We improve the grey wolf algorithm and combine the grey wolf algorithm with the machine learning model. The algorithm constitutes three machine learning hybrid models: GWO-ET, GWO-AdaBoost, and GWO-GBM. Compared with traditional grid tuning, particle swarm optimization (PSO), and genetic algorithm (GA) optimization, the robustness and accuracy of the three optimization algorithms and the machine learning hybrid algorithm on the data set are compared and analyzed. Performance is measured through various performance comparisons and experimental result comparisons. For various building beam component data sets under real fires, the optimization and comparison show that the mean square error (MSE) of the proposed algorithm is extremely small. The results indicate that the GWO machine learning hybrid model is superior to other models and has a smaller prediction error.

Trends in climatically driven extreme growth reductions of Picea abies and Pinus sylvestris in Central Europe.

Extreme tree growth reductions represent events of abrupt forest productivity decline and carbon sequestration reduction. An increase in their magnitude can represent an early warning signal of impending tree mortality. Yet the long-term trends in extreme growth reductions remain largely unknown. We analyzed the trends in the proportion of trees exhibiting extreme growth reductions in two Central-European conifer species-Pinus sylvestris (PISY) and Picea abies (PCAB)-between 1901 and 2018. We used a novel approach for extreme growth reduction quantification by relating their size to their mean recurrence interval. Twenty-eight sites throughout Czechia and Slovakia with 1120 ring width series representing high- and low-elevation forests were inspected for extreme growth reductions with recurrence intervals of 15 and 50years along with their link to climatic drivers. Our results show the greatest growth reductions at low-elevation PCAB sites, indicating high vulnerability of PCAB to drought. The proportions of trees exhibiting extreme growth reductions increased over time at low-elevation PCAB, decreased recently following an abrupt increase in the 1970-1980s at high-elevation PCAB, and showed nonsignificant trends in high- and low-elevation PISY. Climatic drivers of extreme growth reductions, however, shifted over time for all site categories as the proportion of low-temperature-induced extreme growth reductions declined since the 1990s, whereas events caused by drought consistently increased in frequency during the same period. We observed higher growth volatility at the lower range of distribution compared with the upper range margin of PISY and PCAB. This will undoubtedly considerably impact tree growth and vitality as temperatures and incidence of drought in Central Europe are expected to further increase with ongoing climate change.

Prediction of treatment outcome in burning mouth syndrome patients using machine learning based on clinical data

The purpose of this study is to apply a machine learning approach to predict whether patients with burning mouth syndrome (BMS) respond to the initial approach and clonazepam therapy based on clinical data. Among the patients with the primary type of BMS who visited the clinic from 2006 to 2015, those treated with the initial approach of detailed explanation regarding home care instruction and use of oral topical lubricants, or who were prescribed clonazepam for a minimum of 1 month were included in this study. The clinical data and treatment outcomes were collected from medical records. Extreme Gradient-Boosted Decision Trees was used for machine learning algorithms to construct prediction models. Accuracy of the prediction models was evaluated and feature importance calculated. The accuracy of the prediction models for the initial approach and clonazepam therapy was 67.6% and 67.4%, respectively. Aggravating factors and psychological distress were important features in the prediction model for the initial approach, and intensity of symptoms before administration was the important feature in the prediction model for clonazepam therapy. In conclusion, the analysis of treatment outcomes in patients with BMS using a machine learning approach showed meaningful results of clinical applicability.

Equationally extreme trees

We study extreme semilattices whose Hasse diagrams are trees, and prove that the linearly ordered tree of order n has the minimal average number of solutions of 1-variable equations.

Seemingly Unrelated Mixed-Effects Biomass Models for Black Locust in West Poland

Information about tree biomass is important not only in the assessment of wood resources but also in the process of preparing forest management plans, as well as for estimating carbon stocks and their flow in forest ecosystems. The study aimed to develop empirical models for determining the dry mass of the aboveground parts of black locust trees and their components (stem, branches, and leaves). The research was carried out based on data collected in 13 stands (a total of 38 sample trees) of black locust located in western Poland. The model system was developed based on multivariate mixed-effect models using two approaches. In the first approach, biomass components and tree height were defined as dependent variables, while diameter at breast height was used as an independent variable. In the second approach, biomass components and diameter at breast height were dependent variables and tree height was defined as the independent variable. Both approaches enable the fixed-effect and cross-model random-effect prediction of aboveground dry biomass components of black locust. Cross-model random-effect prediction was obtained using additional measurements of two extreme trees, defined as trees characterized by the smallest and largest diameter at breast height in sample plot. This type of prediction is more precise (root mean square error for stem dry biomass for both approaches equals 77.603 and 188.139, respectively) than that of fixed-effects prediction (root mean square error for stem dry biomass for both approaches equals 238.716 and 206.933, respectively). The use of height as an independent variable increases the possibility of the practical application of the proposed solutions using remote data sources.

Abstract P506: Machine Learning Models Improve Prediction of Large Vessel Occlusion and Mechanical Thrombectomy Candidacy in Acute Ischemic Stroke

Background and Purpose: Early identification of large vessel occlusions (LVO) and timely recanalization are paramount in improved clinical outcomes in acute ischemic stroke. Multiple simple stroke scales have good sensitivity, but compromise specificity for predicting LVO. No scale has been shown to predict mechanical thrombectomy (MT) candidacy. Machine learning techniques are being used for predictive modeling in many aspects of stroke care and may have potential in predicting LVO presence and MT candidacy. Methods: 287 acute ischemic stroke patients from July 2018 to July 2019 at Loyola University Medical Center were included. 36 clinical and demographic variables were analyzed using machine learning and statistical algorithms, including logistic regression, extreme gradient boosting, random forest, and decision trees to build models predictive of LVO and MT. The best performing model was compared with prior stroke scales. Results: Random forest based model resulted in the highest classification performance to predict both LVO and MT outcomes with an area under the curve (AUC) of 0.90±0.07 and 0.94±0.04, respectively. When the predictors were reduced to 7, random forest maintained a high AUC for predicting LVO (0.89). When reduced to 10 predictors, the random forest model predicted MT with an AUC = 0.93. Random forest models had excellent sensitivity and specificity of 0.86 and 0.89 for LVO and 0.89 and 0.95 for MT, respectively. The negative predictive value was 0.94 for LVO and 0.98 for MT while the positive predictive value was 0.77 for LVO and 0.79 for MT. With equal sensitivity, the random forest model was favorable to all previous stroke scales. Conclusion: Machine learning utilizing clinical and demographic variables predicts LVO and patient candidacy for MT with a high degree of accuracy. Further validation of this strategy for triage of stroke patients requires prospective and external validation.

Forecasting the Heat Load of Residential Buildings with Heat Metering Based on CEEMDAN-SVR

The energy demand of the district heating system (DHS) occupies an important part in urban energy consumption, which has a great impact on the energy security and environmental protection of a city. With the gradual improvement of people’s economic conditions, different groups of people now have different demands for thermal energy for their comfort. Hence, heat metering has emerged as an imperative for billing purposes and sustainable management of energy consumption. Therefore, forecasting the heat load of buildings with heat metering on the demand side is an important management strategy for DHSs to meet end-users’ needs and maintain energy-saving regulations and safe operation. However, the non-linear and non-stationary characteristics of buildings’ heat load make it difficult to predict consumption patterns accurately, thereby limiting the capacity of the DHS to deliver on its statutory functions satisfactorily. A novel ensemble prediction model is proposed to resolve this problem, which integrates the advantages of Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and support vector regression (SVR), called CEEMDAN-SVR in this paper. The proposed CEEMDAN-SVR algorithm is designed to automatically decompose the intrinsic mode according to the characteristics of heat load data to ensure an accurate representation of heat load patterns on multiple time scales. It will also be useful for developing an accurate prediction model for the buildings’ heat load. In formulating the CEEMDAN-SVR model, the heat load data of three different buildings in Xingtai City were acquired during the heating season of 2019–2020 and employed to conduct detailed comparative analysis with modern algorithms, such as extreme tree regression (ETR), forest tree regression (FTR), gradient boosting regression (GBR), support vector regression (SVR, with linear, poly, radial basis function (RBF) kernel), multi-layer perception (MLP) and EMD-SVR. Experimental results reveal that the performance of the proposed CEEMDAN-SVR model is better than the existing modern algorithms and it is, therefore, more suitable for modeling heat load forecasting.

Bond Risk Premiums with Machine Learning

Abstract We show that machine learning methods, in particular, extreme trees and neural networks (NNs), provide strong statistical evidence in favor of bond return predictability. NN forecasts based on macroeconomic and yield information translate into economic gains that are larger than those obtained using yields alone. Interestingly, the nature of unspanned factors changes along the yield curve: stock- and labor-market-related variables are more relevant for short-term maturities, whereas output and income variables matter more for longer maturities. Finally, NN forecasts correlate with proxies for time-varying risk aversion and uncertainty, lending support to models featuring both channels.

Seemingly Unrelated Mixed-Effects Biomass Models for Young Silver Birch Stands on Post-Agricultural Lands

Secondary succession that occurs on abandoned farmlands is an important source of biomass carbon stocks. Both direct and indirect tree biomass estimation methods are applied on forest lands. Using empirical data from 148 uprooted trees, we developed a seemingly unrelated mixed-effects models system for the young silver birch that grows on post agricultural lands in central Poland. Tree height, biomass of stem, branches, foliage, and roots are used as dependent variables; the diameter at breast height is used as the independent variable. During model elaboration we used restricted cubic spline: 5 knots at the quantiles (0.05, 0.275, 0.5, 0.725, and 0.95) of diameter at breast height provided sufficiently flexible curves for all biomass components. In this study, we demonstrate the use of the model system through cross-model calibration of the biomass component model using tree height measured from 0, 2, 3, and 4 available extreme trees feature in the plot in question. A different number of extreme trees were measured for final model system and our results indicated that for all analyzed components, random-effect predictions are characterized by higher accuracy than fixed-effects predictions.

Deep echo state network: a novel machine learning approach to model dew point temperature using meteorological variables

ABSTRACT The potential of different models – deep echo state network (DeepESN), extreme learning machine (ELM), extra tree (ET), and regression tree (RT) – in estimating dew point temperature by using meteorological variables is investigated. The variables consist of daily records of average air temperature, atmospheric pressure, relative humidity, wind speed, solar radiation, and dew point temperature (T dew) from Seoul and Incheon stations, Republic of Korea. Evaluation of the model performance shows that the models with five and three-input variables yielded better accuracy than the other models in these two stations, respectively. In terms of root-mean-square error, there was significant increase in accuracy when using the DeepESN model compared to the ELM (18%), ET (58%), and RT (64%) models at Seoul station and the ELM (12%), ET (23%), and RT (49%) models at Incheon. The results show that the proposed DeepESN model performed better than the other models in forecasting T dew values.

Mixed-effects generalized height–diameter model for young silver birch stands on post-agricultural lands

The purpose of creating regression equations is often to predict unmeasured features based upon more easily obtainable ones. Species-specific height–diameter (H–D) models of trees are an example of this situation and can be defined as either simple or generalized. Simple H–D models express height as a function of tree diameter at the breast height. They are easily applicable without additional measurement but do not take properly into account the variability in H-D relationship between stands. Meanwhile, generalized models also include stand-level predictors. The H-D data sets are often characterized by a grouped structure. The mixed-effects modeling approach is a mainstream method employed for these types of forestry data. In this study, we created a mixed-effects generalized H–D model for young silver birch stands on post-agricultural lands in central Poland. This model was chosen from among 11 simple nonlinear models based on the goodness of fit and residual behavior. We accounted for two stand-level predictors that did not require additional measurements beyond tree diameter at the breast height: quadratic mean diameter at the breast height and basal area. Fixed- and random-effect predictions were then calculated to illustrate that increases in the number of measured trees improves height predictions. Moreover, the gain in predictive power is the largest if extreme trees (i.e., from the extrema of the diameter range) are used for random-effect prediction.

Data Driven Determination of Reaction Conditions in Oxidative Coupling of Methane via Machine Learning

AbstractThe challenge in catalytic reactions lies within its complexity coming from high dimensional experimental factors. In order to solve such complexity, machine learning is implemented to treat experimental conditions in high dimensions. Oxidative coupling of methane, methane to C2 compounds (ethylene and ethane), is chosen as the prototype reaction where 156 data consisting of various experimental conditions is prepared. Machine learning reveals that the relationship between experimental conditions and C2 yield is non‐linear matter. In particular, extreme tree regression is found to accurately reproduce the experimental data. In addition, machine learning predictions can be a good indicator for designing experiments. Thus, machine learning can be a powerful approach towards understanding and determining experimental conditions in high dimension.

Slow Growth Leads to Longevity in Temperate Hardwoods

Discovering, studying, and preserving old trees is a top priority for conservation biology. We used tree-ring data from a high-mountain old-growth beech forest to reconstruct long-term growth patterns in trees of maximum longevity for temperate hardwoods. In these stands, individual growth history is highly variable; one tree can take from one to seven centuries to reach a large size (>60 cm dbh). Slow but overall increasing long-term growth was found to be a prerequisite for extreme tree longevity. Growth suppression in early stages of life, together with limiting climatic conditions, also contributed to reaching maximum stem ages in these Mediterranean mountain environments. Photo credit: Gianluca Piovesan. Photo credit: Gianluca Piovesan. Photo credit: Gianluca Piovesan. Photo credit: left and center photos, Gianluca Piovesan; right photo, Vittoria Marchianò. Photo credit: left and center photos, Vittoria Marchianò; right photo, Giuseppe De Vivo. These photographs illustrate the article: “Lessons from the wild: Slow but increasing long-term growth allows for maximum longevity in European beech” by Gianluca Piovesan, Franco Biondi, Michele Baliva, Giuseppe De Vivo, Vittoria Marchianò, Aldo Schettino, and Alfredo Di Filippo published in Ecology. https://doi.org/10.1002/ecy.2737

Bond Risk Premia with Machine Learning

We show that machine learning methods, in particular extreme trees and neural networks (NNs), provide strong statistical evidence in favor of bond return predictability. NN forecasts based on macroeconomic and yield information translate into economic gains that are larger than those obtained using yields alone. Interestingly, the nature of unspanned factors changes along the yield curve: stock and labor market related variables are more relevant for short-term maturities, whereas output and income variables matter more for longer maturities. Finally, NN forecasts correlate with proxies for time-varying risk aversion and uncertainty, lending support to models featuring both of these channels.

Analysis of Electric Load Forecasting using Artificial Intelligence

Electricity demand forecasts are extremely important for energy suppliers and other actors in production, transmission, distribution and energy markets. Accurate models for predicting the load of electricity are critical to the operation and planning of a service company. load forecasts are extremely important for energy suppliers and other actors in production, transmission, distribution and energy markets. short-term load forecasts play an important role in the operation of power systems to ensure an immediate balance between energy production and demand. The accuracy of the prediction generated by the artificial intelligence has several factors, including, but not limited to, what is used to form the network algorithm, how much and which type of data are used in the network’s training set. In this research the best combination has been investigated of these factors to decrease performance parameters to give the best forecast possible. Artificial intelligence techniques have gained importance in reducing estimation errors. Artificial neural network, Extreme Learning Machine and Decision tree such as LSBOOST and RF are among these artificial intelligence techniques. That are used in this research work for performance analysis. In this work, performance evaluation metrics such as MSE, RMSE, MAE and MAPE values are analysed and it is concluded that Random forest decision tree forecasting algorithm gives better performance of forecasting as compared to other artificial intelligence algorithms for 24 hours load forecasting as well as for 7 day load forecasting.

Tree ring‐based metrics for assessing old‐growth forest naturalness

SummaryOld‐growth studies commonly emphasize structural and age conditions, selecting proxy indicators of long‐term ecological processes. Transition dynamics from mature to old‐growth status reveal how natural legacies are progressively accumulated in forests after major disturbances, including human ones. In late‐successional, multi‐aged forests, the chronosequential ranking of developmental stages is a difficult task, as stand age provides little information, and time since last stand‐replacing disturbance cannot be easily determined.Canopy age features, disturbance/suppression history and growth trajectories were reconstructed from ring‐width series of canopy trees in a network of 19 old‐growth and managed European beech forests in the eastern Alps and central Apennines. A set of tree‐ring metrics able to describe the intensity and time distribution of biological and ecological processes (e.g. understorey suppression, canopy accession age) were used to describe the advancement of old‐growth status and compared to established metrics of forest structure.Tree‐ring metrics were site dependent, as biogeoclimate affects turnover rates and constrains the onset and recovery rate of old‐growth attributes. Under the same environmental conditions (high‐mountain, limestone‐bedrock beech forests), values of the best indicators (number/duration of growth suppression phases; synchronicity of first release; maximum and range of canopy tree age; canopy accession age of the slowest‐growing trees) increased monotonically (two to five times) from managed to secondary and primary old‐growth forests. Trees in well‐conserved primary old‐growth forests experienced several and long suppressions, showing the highest complexity in recruitment history, canopy accession and growth trajectories. The best tree‐ring metrics, condensed in a Naturalness Score to provide a synthetic functional ranking of forests, varied coherently with structural complexity, which represented stand dynamics more closely than biomass‐related metrics.Synthesis and applications. We propose a synthetic ranking of forest functional naturalness based on the ecological processes experienced by trees. This ranking helps to overcome the limitations associated with the use of arbitrary size‐ or age‐related thresholds of old‐growth status and provides a functional approach to establish chronosequences in ecological studies. The quantitative description of complex processes underpinning the unique biological and ecological features (e.g. extreme tree longevity) found in primary old‐growth forests enhances their irreplaceable value in nature conservation. The proposed framework of tree‐ring indicators describes functional traits tightly related to forest naturalness and may thus become a tool to identify and protect old‐growth forests, benchmark the impact of silvicultural practices, prescribe targets or evaluate the effectiveness of restoration programmes.

Applications of Hybrid Extreme Rotation Forests for image segmentation

This paper introduces the Hybrid Extreme Rotation Forest HERF classifier describing two succesful applications in the image segmentation domain. The HERF is an ensemble of classifiers composed of Extreme Learning Machines ELM and Decision Trees. Training of the HERF includes optimal rotation of random partitions of the feature set aimed to increase diversity. The first application is the segmentation of 3D Computed Tomography Angiography CTA following an Active Learning AL strategy for the optimal sample selection to minimize the number of data samples needed to obtain a required accuracy degree. AL is pertinent for interactive learning processes where a human operator is required to select training samples to enhance the classifier in an iterative process, therefore labeling samples for training may be a time consuming and expensive process. CTA image segmentation is one of such processes, due to the variability in CTA images which hinders the generalization of classifiers trained on one dataset to new datasets. Following an AL strategy, the human operator is presented with a visual selection of pixels whose labeling would be most informative for the classifier. After adding those labeled pixels to the training data, the classifier is retrained. This iteration is repeated until image segmentation quality meets the required level. The approach is applied to the segmentation of the thrombus in CTA imaging of Abdominal Aortic Aneurysm AAA patients, showing that the structures of interest can be accurately segmented after a few iterations using a small data sample. The second application is a new semisupervised segmentation algorithm for hyperspectral images. The algorithm steps are: 1 supervised training an initial classifier from a small balanced training set, 2 clustering of the image pixels, by a k-means algorithm 3 adding unlabeled pixels to the original trainning data set according to the spatial neighborhood and the cluster membership, 4 supervised training of the classifier with the enriched training data set, 6 classification of the hyperspectral image 4 spatial regularization of classification results consisting in selecting the most frequent class in each pixel neighborhood. Results on two well known benchmarking hyperspectral images improve over state of the art algorithms.

Phylogenetically Clustered Extinction Risks Do Not Substantially Prune the Tree of Life

Anthropogenic activities have increased the rate of biological extinction many-fold. Recent empirical studies suggest that projected extinction may lead to extensive loss to the Tree of Life, much more than if extinction were random. One suggested cause is that extinction risk is heritable (phylogenetically patterned), such that entire higher groups will be lost. We show here with simulation that phylogenetically clustered extinction risks are necessary but not sufficient for the extensive loss of phylogenetic diversity (PD) compared to random extinction. We simulated Yule trees and evolved extinction risks at various levels of heritability (measured using Pagel's ). At most levels of heritability ( in range of 0 to 10), mean values of extinction risk (range 0.25 to 0.75), tree sizes (64 to 128 tips), tree balance and temporal heterogeneity of diversification rates (Yule and coalescent trees), extinction risks do not substantially increase the loss of PD in these trees when compared to random extinction. The maximum loss of PD (20% above random) was only associated with the combination of extremely excessive values of phylogenetic signal, high mean species' extinction probabilities, and extreme (coalescent) tree shapes. Interestingly, we also observed a decline in the rate of increase in the loss of PD at high phylogenetic clustering of extinction risks. Our results suggest that the interplay between various aspects of tree shape and a predisposition of higher extinction risks in species-poor clades is required to explain the substantial pruning of the Tree of Life.

Generalized Maxwell formula for the length of a minimal tree with a given topology

The classic Maxwell formula calculates the length of a planar locally minimal binary tree in terms of coordinates of its boundary vertices and directions of incoming edges. However, if an extreme tree with a given topology and a boundary has degenerate edges, then the classic Maxwell formula cannot be applied directly, to calculate the length of the extreme tree in this case it is necessary to know which edges are degenerate. In this paper we generalize the Maxwell formula to arbitrary extreme trees in a Euclidean space of arbitrary dimension. Now to calculate the length of such a tree, there is no need to know either what edges are degenerate, or the directions of nondegenerate boundary edges. The answer is the maximum of some special linear function on the corresponding compact convex subset of the Euclidean space coinciding with the intersection of some cylinders.

On the number of independent sets in the trees of a fixed diameter

A few lower bounds are obtained for the number of independent sets in the trees of diameter 6, 7, 8, and 9. Some characteristic is given of the extreme trees of an arbitrary diameter.