Sparse Tree Research Articles

Can rewilding with giant tortoises increase woody habitat and limit fire across Madagascar's grasslands?

Societal Impact StatementTo better understand the impact of human settlement on vegetation distribution and disturbance regimes in Madagascar, the 2021 Malagasy Grassy Biomes Workshop called for an examination of the role of extinct megafauna in determining the historic distribution of habitat mosaics across regions that are today dominated by vast treeless grasslands. A systematic literature review facilitated the reconstruction and evaluation of giant tortoise habitat, feeding habitats, diets and Aldabrachelys tortoise impacts on disturbance regimes. In conclusion, reintroduction of mixed‐feeding Aldabrachelys (the diets of which were often dominated by C3 woody material, but never by C4 grasses, across its range), will likely impact plant species composition, select a tree‐rich habitat mosaic, reduce fire frequency, re‐establish lost ecological functions and improve landscape productivity at a range of scales, including rangeland palatability, benefitting both biodiversity and human wellbeing.SummaryGrasslands with little tree cover today comprise 80% of Madagascar's habitat. Determining their extent at human settlement can guide ecological restoration and enhance human well‐being, so the 2021 Malagasy Grassy Biomes Workshop identified the role of extinct megafauna in determining habitat as a critical knowledge gap. Using a systematic literature review, combined with extracted datasets, we address this, examining anticipated habitat selection by giant tortoises following reintroduction to Madagascar (where the Aldabran giant tortoise, Aldabrachelys gigantea, provides ecological functions lost when A. abrupta and A. grandidieri went extinct). When comparing current and historical tortoise selection of habitat across the Mascarenes and Aldabra with contemporary Malagasy habitat, areas in Madagascar where giant tortoises historically ranged, today have a significantly different habitat composition to the forested habitat that supported giant tortoises on other islands. Dietary 13C isotope ratios show that Malagasy Aldabrachelys and Mascarene tortoises were mixed feeders, with diets often dominated by C3 woody intake, but never by C4 grasses. Across systems, giant tortoises required and selected, tree‐rich habitat mosaics, different to current pastoralist fire‐selected Malagasy grasslands characterized by sparse tree cover. Furthermore, Aldabran Aldabrachelys tortoise turf, restricted to small areas (large tracts of unshaded vegetation present physiological challenges to Aldabrachelys' survival), is compositionally different to Malagasy and African obligate C4 grazing lawns. Ecological, palaeoecological, geomorphological and molecular evidence support a lost Malagasy habitat mosaic where hippo and tortoise diets were C3‐dominated, because they inhabited closed‐canopy systems, with abutting open‐canopy areas harbouring endemic‐rich, C4 grassy understories and limited grasslands. The review suggests that rewilding with A. gigantea will help restore ecological functions, productivity and landscape‐scale degradation lost through cattle‐based pastoralism, re‐establish tree‐rich habitat mosaics, and mitigate against frequent bushfires, benefiting biodiversity and humans at multiple scales.

Does chronic nitrogen deposition have effects on grass physiology of natural habitats?

Anthropogenic activities increase nitrogen deposition, which can have major consequences for biodiversity and ecosystem function. We experimentally investigated how a community of grass species would respond to three levels of nitrogen enrichment, considering their biomass and efficiency of photosynthesis. The experiments were conducted in an open savanna area with sparse trees (cerrado sensu stricto). Fifteen plots were divided considering high nitrogen supplementation, low nitrogen supplementation, and control (without nitrogen application) as treatments. The nitrogen dosages applied in each treatment were based on deposition estimates for 2050 in the Cerrado biome. Three native species had their photosynthetic metabolism evaluated 45 days after the last supplementation. Except for Echinolaena inflexa, all the 16 species found in the area expressed C4 metabolism. Echinolaena inflexa and Loudetiopsis chrysothrix showed an increase in leaf nitrogen with supplementation, with differences in chloroplastidic pigment contents. Differently from L. chrysothrix which increased its chlorophyll content, E. inflexa showed a decrease in chlorophyll and carotenoid contents under N supplementation. E. inflexa is a small species and its leaf expansion under higher contents of nitrogen suggests that this species was affected by shading caused by other community species that increased their biomass. Supplemental nitrogen did not affect the efficiency of photosynthesis of these species or severely change the grass community biomass. However, we noted only a tendency for higher biomass and biomass per tiller for Urochloa decumbens, which suggests a concern about exotic species' performance under conditions of greater availability of nitrogen.

Voltage Stability Constrained Operation Optimization: An Ensemble Sparse Oblique Regression Tree Method

Voltage stability remains a main challenge for high renewable penetrated power systems. In future power systems, the dispatch needs to consider the voltage stability margin. However, considering static and transient voltage stability is challenging because a complex relationship exists between voltage stability and operation state variables. This paper proposes an ensemble sparse oblique regression tree method for voltage stability constrained operation optimization. First, we train multiple oblique regression trees on voltage stability simulation datasets. The trained trees are then grouped into ensemble using a boosting method to extract accurate and understandable voltage stability rules. Finally, the rules are embedded as mixed-integer linear programming constraints in a linear optimal power flow model. Validation on the IEEE-14 case shows the interpretability and efficiency of the proposed algorithm. Case studies on the Qinghai power grid and IEEE-118 further show that the proposed strategy can effectively improve the voltage stability margin of optimized operation states.

Influence of obstacles on urban canyon ventilation and air pollutant concentration: An experimental assessment

Air pollution in cities, intensified by vehicular traffic emissions and reduced ventilation, poses a significant health risk. Obstacles, like solid or vegetation barriers, are being considered as strategies to reduce pollution exposure for pedestrians and nearby residents in street canyons. This study utilises wind tunnel experiments, to simulate a typical urban canyon with street intersections on both sides, in a 1:200 scale and an H/W=0.5, positioned perpendicular to the free stream wind flow. A passive scalar representing a vehicular pollutant is released along the length of the street canyon. Concentration measurements inside the canyon are performed to determine the effect of parked cars, boundary walls, hedges and trees on pollutant concentration exposure for pedestrians on the sidewalk. Results show that one circulating vortex is generated within the canyon, driving the pollutant to accumulate along the leeward (upwind) wall. Tightly parked cars, boundary walls and hedges, placed along the sidewalk near the leeward wall, can reduce pedestrian pollutant exposure by 15 %, 23 % and 11 % respectively along this sidewalk. Attributes such as obstacle height, surface roughness and porosity play a key role in their performance. However, trees, when placed in the same area, increase pedestrian pollutant exposure by 51 % and 17 % under dense and sparse tree arrangements, respectively. While a broader analysis that considers the variability of vegetation attributes (e.g., porosity, stand density) is desirable, this study remains crucial for validating numerical simulations and suggesting optimal urban measures to reduce pollution exposure for citizens.

Environmental evolution of western Tierra del Fuego (∼54°S) since ice-free conditions and its zonal/hemispheric implications

By virtue of its position adjacent to the Drake Passage, Tierra del Fuego in South America allows examining the vegetation and environmental history of the southernmost continental landmass outside Antarctica, and the evolution of the Southern Westerly Winds-Southern Ocean (SWW–SO) coupled system since the Last Glacial Maximum (LGM). For that purpose, we studied sediment cores from Lago Charquito, a small closed-basin lake in central-west Tierra del Fuego with a continuous lacustrine record since ∼17.3 ka. Ice-free conditions at the site imply a ∼70 km retreat of the Bahía Inútil glacier lobe from its LGM position during a ∼800-year interval, a trend that continued until its disappearance ∼100 km upstream from L. Charquito, ∼800 years later. Our palynological data show an open landscape dominated by cold-tolerant shrubs and herbs between ∼17.3–12.9 ka, with increases in precipitation of SWW origin at ∼16.3 ka, ∼14.7 ka, between ∼8.7–7.6 ka, and after ∼6.8 ka. Warming at ∼12.9 ka initiated an abrupt afforestation trend that stalled during the early Holocene (∼12–8.7 ka) owing to a precipitation decline and wildfires, and later resumed in response to invigorated SWW. We hypothesize that sparse Nothofagus tree populations inhabited the periphery of the Patagonian Ice Sheet (PIS) during the LGM and migrated toward the Andes contemporaneous with glacier recession as temperature rose during the Last Glacial Termination (T1). We posit that besides establishing topographic and climatic barriers for land biota, the PIS enabled the connectivity of cold-tolerant hygrophilous plant populations along a humid fringe adjacent to its land-based perimeter, despite the presumably dry conditions downwind from the eastern PIS margin. Our results suggest that southward shifts or expansion of the SWW toward or beyond Tierra del Fuego enhanced upwelling and ventilation of deep waters in the SO, northward shifts or weakening had the opposite effect. We observe that the time evolution of atmospheric CO2 concentrations, high-latitude air and sea-surface temperatures, and sea level during T1 fall short in explaining the timing and abruptness of the Bahía Inútil glacier lobe collapse, and quite possibly multiple other glacier lobes from the PIS.

An Efficient Two-Stage SPARC Decoder for Massive MIMO Unsourced Random Access

In this paper, we study a concatenate coding scheme based on sparse regression code (SPARC) and tree code for unsourced random access in massive multiple-input and multiple-output systems. Our focus is concentrated on efficient decoding for the inner SPARC with practical concerns. A two-stage method is proposed to achieve near-optimal performance while maintaining low computational complexity. Specifically, a one-step thresholding-based algorithm is first used for reducing large dimensions of the SPARC decoding, after which a relaxed maximum-likelihood estimator is employed for refinement. Adequate simulation results are provided to validate the near-optimal performance and the low computational complexity. Besides, for covariance-based sparse recovery method, theoretical analyses are given to characterize the upper bound of the number of active users supported when convex relaxation is considered, and the probability of successful dimension reduction by the one-step thresholding-based algorithm.

KARAKTERISTIK PERAIRAN MANGROVE PADA KERAPATAN YANG BERBEDA DI DESA EYAT MAYANG KABUPATEN LOMBOK BARAT

Different characteristics of mangrove habitat will have different water conditions. The purpose of this study was to determine the characteristics of mangrove habitat waters at different mangrove densities in Eyat Mayang Village, Lombok Barat Regency. Observation of mangrove density and characteristics of the waters was carried out in each sample plot using the grid line method. The tree parameters observed included the type and number of trees in the 84 sample plots. Parameters of mangrove habitat waters observed included temperature, pH, salinity, dissolved oxygen (DO) and mud thickness. Mangrove density is divided into 3 criteria based on the Decree of the Minister of Environment No. 201 of 2004. Furthermore, statistical analysis was carried out to assess differences in water characteristics at 3 mangrove densities using one-way analysis of variance with a level of 5% (significance 0.05) with SigmaPlot 11.0 software. The results showed that the sparse, medium and dense densities had an average temperature range of 29 ºC, a pH value of 6.8 – 7, a salinity in the range of 35‰, DO 4.8 – 5 mg/l and a mud thickness of 72 – 91 cm . The mangrove area of Eyat Mayang Village has 77.1% sparse tree density, 20% medium tree density and 2.9% dense tree density. The results of statistical analysis using the ANOVA test showed that there were no significant differences in the parameters of temperature, salinity, pH, DO and mud thickness at 3 different density of mangrove trees.

The analysis and application of Prim algorithm, Kruskal algorithm, Boruvka algorithm

Minimum spanning tree has many applications in real life. For example, the government needs to build roads between many cities. Therefore, it is necessary to find the plan with the shortest path to save the cost. The problem is essentially generating a minimal spanning tree, and it require a suitable algorithm to find the minimum spanning tree. In this paper, the author analyzes the structure and time complexity of the Prim algorithm, the Kruskal algorithm and the Boruvka algorithm. Through this research, the author finds Prim algorithm is suitable for dense graphs. The Kruskal algorithm can generate the minimum spanning tree in sparse tree. And the Boruvka algorithm is suitable for graphs that have some special characters. Based on the above conclusions, the author gives some suggestions for urban highway network planning.

Surviving trees are key elements in the fate of ectomycorrhizal community after severe bark-beetle forest disturbance.

Bark beetle disturbances are a critical event in the life cycle of Norway spruce forests. However, our knowledge of their effects on ectomycorrhizal fungi (EMF), which play a key role in forest productivity and nutrient cycling, is still incomplete. Special attention has been paid to the dynamics and diversity of EMF communities in managed forests, but studies dealing with disturbed natural stands are underrepresented. We conducted a study in an unmanaged natural spruce forest in the Bohemian Forest, (Czech Republic) which suffered severe forest dieback caused by bark beetle. Approximately a decade after the disturbance, the character of the forest structure in the study area (∼60 ha, 41 study plots) ranged from sites with open canopy and sparse tree cover to areas with dense spruce regeneration to patches of closed-canopy forest. We found that relative EMF abundance in soils was positively related to surviving tree and regeneration density. The number of surviving trees also positively affected species EMF richness and tended to support preservation of late-successional EMF species. Our results suggest that trees that survive bark beetle disturbance are key for the fate of the EMF community in natural forests.

Accuracy of Manual Snow Sampling, Depending on the Sampler’s Cross-Section—A Comparative Study

Snow sampling, either by inserting a tube through the entire snowpack or by taking samples from the vertical profile, is widely applied to measure the snow depth, density, and snow water equivalent (SWE). A comparative study of snow-sampling methods was carried out on 24 March 2022 in Sodankylä, Finland. Six groups from five countries (Estonia, Finland, Greenland, Iceland, and Sweden) participated, using 12 different snow samplers, including 9 bulk tube samplers and 3 density cutters. The cross-sectional area of the SWE samplers varied from 11 to 100 cm2, while tube length varied from 30 cm to 100 cm. The cross-sectional area of the density profile cutters varied from 100 cm2 to 200 cm2 and the vertical sampling step varied from 5 cm to 10 cm. The samples were taken from snow pits in 55–65-centimeter-deep snow cover in a flat area with sparse pine trees, with the pits at a maximum distance of 10 m from each other. Each tube sampling series consisted of 3–10 vertical columns to ensure statistical validation. The snowpack was relatively soft, with two moderately hard crust layers. The density recorded in the tube sample measurements varied from 218 to 265 kgm−3. The measurement results of SWE, however, varied depending on the sampling equipment used, ranging from 148 to 180 kgm−2, with two outliers of 77 and 106 kgm−2, both with 11 cm2 samplers.

Optimal Sparse Regression Trees.

Regression trees are one of the oldest forms of AI models, and their predictions can be made without a calculator, which makes them broadly useful, particularly for high-stakes applications. Within the large literature on regression trees, there has been little effort towards full provable optimization, mainly due to the computational hardness of the problem. This work proposes a dynamic-programming-with-bounds approach to the construction of provably-optimal sparse regression trees. We leverage a novel lower bound based on an optimal solution to the k-Means clustering algorithm on one dimensional data. We are often able to find optimal sparse trees in seconds, even for challenging datasets that involve large numbers of samples and highly-correlated features.

Measuring the relationships between various urban green spaces and local climate zones

Urban green spaces (UGS) improve living conditions in cities by mitigating the Urban Heat Island effect. While the cooling effect of UGS seems unequivocal, the relationship between the types of UGS and types of residential areas has not yet been well explored. In this study, we systematically analysed the cooling effect of 71 UGS in Prague, a central European city, on residential areas within 400 m of the UGS. The UGS are classified according to their spatial characteristics (size, shape, and tree density), and the residential areas according to three Local Climate Zones (LCZ 2, 5, 6) typical for European cities. The cooling effect is evaluated using a regression model of the Land Surface Temperature (LST) in residential zones according to the LCZ type and distance from the various UGS. The results show that compact UGS of 10–25 ha with dense trees have the most pronounced cooling effect. This type of UGS was associated with a mean decrease in LST within 400 m of 2.3 °C compared to the least effective UGS type (long with sparse trees) across LCZs. The results of the presented study can be applied in urban planning and urban design to improve microclimates in cities.

Monitoring forest dynamics in Africa during 2000–2020 using a remotely sensed fractional tree cover dataset

ABSTRACT Africa is covered with extensive woodlands, savannas and rainforests. The tree cover of these biomes has been undergoing substantial changes in recent decades. However, the dynamics of forests in Africa are currently unclear, particularly in the woodlands and savanna areas covered by sparse trees. Here, we assessed the spatio-temporal trend of African forests from 2000 to 2020, using a 250-m resolution fractional tree cover product that can capture the variation of forest density in the widespread mixed vegetation landscapes of the continent. The tree cover trends, interannual change and hotspots of forest gain and loss were evaluated. Results showed that the African forest area increased at a rate of 3.59 million ha/year over the study period, reaching 589 million ha in 2020. Considerable forest gain and loss both occurred in Africa. The net change rate in woodlands’ forest area was the fastest (2.28 million ha/year), followed by rainforests (0.80 million ha/year) and savannas (0.34 million ha/year). Hotspots of forest gain were concentrated in the north belt of woodlands and savannas, while forest loss primarily clustered in East and South Africa. This work would help African countries to monitor forest change and promote forest management to achieve the Sustainable Development Goals.

Restoring functional forested peatlands by combining ditch-blocking and partial cutting: An amphibian perspective

Artificial drainage for forestry has caused major degradation of temperate peatland ecosystems. Partial restoration of those ecosystems is now widely considered, but feasible targets and effectiveness of restoration practices are not well established. In this study, we analyse a forested peatland restoration experiment in Estonia from a functional perspective of amphibian populations. In 2014–2020, we surveyed a before-after-control-impact experiment that combined two habitat manipulations: partial cuttings for restoring sparse tree canopy cover, and ditch blocking with soil filling to restore the water table. We monitored the responses in the aquatic habitats and breeding of brown frogs (moor frog Rana arvalis and common frog R. temporaria). Along with the expected habitat changes (ditch blocking quadrupled water depths and tree cutting reduced the shade), we found that ditch blocking restored the soil-type specific water acidity levels that the draining had homogenized. Partial ditch blocking reduced frog breeding activity in the first year, but it recovered and increased later on, specifically on sapric Histosols and supported by temporary wetlands, floods and pools created. However, the number of offspring did not clearly respond to the restoration and instead varied much among years. Our results indicate that, in drained forested wetlands: (i) near-natural aquatic habitats are a distinct restoration target and several years of post-restoration monitoring are needed to confirm the development of ecological integrity; (ii) increased breeding activity of amphibians may not result in improved recruitment; and thus, (iii) integrative wetland restoration techniques could benefit from an “amphibian perspective” of sustaining their whole life cycles and the food webs they support.

Field-Layer Vegetation and Water Table Level as a Proxy of CO2 Exchange in the West Siberian Boreal Bog

The Mukhrino field station has participated in the national project on the inventory of carbon fluxes and pools in the terrestrial ecosystems of Russia since 2022. The development of a network of measurements of CO2 fluxes and phytomass covered six types of bog ecosystems typical to Western Siberia. The gross ecosystem exchange (GEE) of the field-layer vegetation (medians for the period from the end of May to the end of July, mgC m−2 h−1; see errors in Results section) decreased in series: Sphagnum bog with sparse low pine trees (“Open bog”), ridges in ridge-hollow patterned bogs (“Ridge”), pine-dwarf shrub-Sphagnum bog (“Tall ryam”), hollows in patterned bogs (“S.hollow”, “E.hollow”) and pine-dwarf shrub-Sphagnum bog (“Ryam”): −220, −200, −125, −120, −109 and −86, respectively. Ecosystem respiration (Reco) here was 106, 106, 182, 55, 97 and 136. The aboveground and belowground phytomass of mosses in this series varied between 368 ± 106–472 ± 184 and 2484 ± 517–6041 ± 2079 g/m2, respectively: the aboveground phytomass of vascular plants and plant litter—15 ± 7–128 ± 95 and 10 ± 6–128 ± 43, respectively. According to the results of mathematical modeling, the best proxy for GEE, in addition to photosynthetically active radiation and soil surface temperature, was the aboveground phytomass of vascular plants (PhV), and for Reco—PhV and the mass of the plant litter of vascular plants.

Overlay

Overlay

Phytocenotic conditions and structure of the local population for Betula humilis Schrank near Zhornyska village (Ukrainian Roztochchia)

hytocoenotic conditions and structure of the local population for Betula humilis in the north-eastern environs of the village Zhornyska, Yavoriv district, Lviv region, on the territory of the Roztochchia International Biosphere Reserve have been studied. Four coenopopulations (CP) of the species are confined to secondary plant communities of sicaciogenic and pyrogenic successions with different sod levels of the peat soil, as well as missing or unequally expressed tree layer. The least transformed communities with the dominance of B. humilis (40-80% p.c.) in the shrub layer and no stands in terms of floristic composition and spatial structure correspond to the derived mesohygrophytic and hygromesophytic variants of the association Betuletum humilis Steffen 1931. The most transformed areas with sparse tree layer are represented by pioneer forest communities - birch woodlands. Seventy-five species of higher plants (seventy vascular plants and five bryophytes) were found in the floristic composition of phytocenoses. The minimal participation of helophytes and the presence of invasive species indicate a high level of mesophytization and synanthropization of modern phytocenoses. Within the locality B. humilis forms different in area and density of formation shoots (turions) of the colony-curtain mainly from semi-autonomous subpartial bushes 0.3-2.2 m tall. The largest curtains (up to 300 m2) with a high density of living formation shoots (up to 293 pcs/m2) are confined to open well-lit areas. The age of old skeletal axis, not exceeding 9 years was determined and the annual growth of shoots has been calculated. The parameters of seed productivity (SP) of the local population of B. humilis vary within: potential SP - 89.48–109.24 seed germs, actual SP - 80.84–100.4 seeds. The coefficient of seed productivity is quite high (86.69-92.16%) for all coenopopulations. Based on the results of research and analysis of the literature the life form of B. humilis, which belongs to vegetatively mobile hypogeogenic-geoxyl deciduous shrubs that form colonies or clones of (semi) autonomous (sub) partial (daughter) shrubs and are characterized by monopodial-sympodial shoots growth has been clarified. The big life cycle of the species characteristic of shrubs consists of development cycles for systems shoots formation (main skeletal axes), which are successively replaced, and lasts for decades. The development cycle of the system formation shoot covers the periods of intensive growth (appearance of formation shoot from an additional or dormant bud in the first year), relative stabilization of growth processes with the formation of first and higher-order branching shoots (from second to tenth years) and ageing and death of shoots (after about ten years of age).

Tree Branch Skeleton Extraction from Drone-Based Photogrammetric Point Cloud

Calculating the complex 3D traits of trees such as branch structure using drones/unmanned aerial vehicles (UAVs) with onboard RGB cameras is challenging because extracting branch skeletons from such image-generated sparse point clouds remains difficult. This paper proposes a skeleton extraction algorithm for the sparse point cloud generated by UAV RGB images with photogrammetry. We conducted a comparison experiment by flying a UAV from two altitudes (50 m and 20 m) above a university orchard with several fruit tree species and developed three metrics, namely the F1-score of bifurcation point (FBP), the F1-score of end point (FEP), and the Hausdorff distance (HD) to evaluate the performance of the proposed algorithm. The results show that the average values of FBP, FEP, and HD for the point cloud of fruit tree branches collected at 50 m altitude were 64.15%, 69.94%, and 0.0699, respectively, and those at 20 m were 83.24%, 84.66%, and 0.0474, respectively. This paper provides a branch skeleton extraction method for low-cost 3D digital management of orchards, which can effectively extract the main skeleton from the sparse fruit tree branch point cloud, can assist in analyzing the growth state of different types of fruit trees, and has certain practical application value in the management of orchards.

Sparse oblique decision trees: a tool to understand and manipulate neural net features

The widespread deployment of deep nets in practical applications has lead to a growing desire to understand how and why such black-box methods perform prediction. Much work has focused on understanding what part of the input pattern (an image, say) is responsible for a particular class being predicted, and how the input may be manipulated to predict a different class. We focus instead on understanding which of the internal features computed by the neural net are responsible for a particular class. We achieve this by mimicking part of the neural net with an oblique decision tree having sparse weight vectors at the decision nodes. Using the recently proposed Tree Alternating Optimization (TAO) algorithm, we are able to learn trees that are both highly accurate and interpretable. Such trees can faithfully mimic the part of the neural net they replaced, and hence they can provide insights into the deep net black box. Further, we show we can easily manipulate the neural net features in order to make the net predict, or not predict, a given class, thus showing that it is possible to carry out adversarial attacks at the level of the features. These insights and manipulations apply globally to the entire training and test set, not just at a local (single-instance) level. We demonstrate this robustly in the MNIST and ImageNet datasets with LeNet5 and VGG networks.

Fuzzy Gaussian Decision Tree

The Decision Tree algorithm is one of the first machine learning algorithms developed. It is used both as a standalone model and as an ensemble of many cooperating trees like Random Forest, AdaBoost, Gradient Boosted Trees, or XGBoost. In this work, a new version of the Decision Tree was developed for classifying real-world signals using Gaussian distribution functions and a fuzzy decision process. The research was carried out on Power Quality Classification Dataset hosted on the platform kaggle.com for accessing the algorithm’s classification. The proposed algorithm modification produces a sparse tree of multidimensional Gaussian kernels performing fuzzy, proximity-based division of solution space instead of a typical Decision Tree performing definite space restrictions. The machine learning model can cluster samples based on the common pattern and evaluate the input’s similarity in a fuzzified fashion. The studies were conducted for the prediction of 6 classes of current. The proposed modification achieved the best accuracy and F1 score compared to the default Decision Tree and other machine learning algorithms.