Multiple Trees Research Articles

Terraces in species tree inference from gene trees

A terrace in a phylogenetic tree space is a region where all trees contain the same set of subtrees, due to certain patterns of missing data among the taxa sampled, resulting in an identical optimality score for a given data set. This was first investigated in the context of phylogenetic tree estimation from sequence alignments using maximum likelihood (ML) and maximum parsimony (MP). It was later extended to the species tree inference problem from a collection of gene trees, where a set of equally optimal species trees was referred to as a “pseudo” species tree terrace which does not consider the topological proximity of the trees in terms of the induced subtrees resulting from certain patterns of missing data. In this study, we mathematically characterize species tree terraces and investigate the mathematical properties and conditions that lead multiple species trees to induce/display an identical set of locus-specific subtrees owing to missing data. We report that species tree terraces are agnostic to gene tree heterogeneity. Therefore, we introduce and characterize a special type of gene tree topology-aware terrace which we call “peak terrace”. Moreover, we empirically investigated various challenges and opportunities related to species tree terraces through extensive empirical studies using simulated and real biological data. We demonstrate the prevalence of species tree terraces and the resulting ambiguity created for tree search algorithms. Remarkably, our findings indicate that the identification of terraces could potentially lead to advances that enhance the accuracy of summary methods and provide reasonably accurate branch support.

Seasonal vertical accretion in different hydro-geomorphic environments of Indian Sundarban: Assessing the role of mangroves in mitigating sea level rise impacts

The Indian Sundarban, renowned for its unique hydro-geomorphic diversity, faces rising threats from sea level rise, necessitating an examination of the spatio-temporal variation of vertical accretion and the role played by mangroves in mitigating these impacts. As such, fifteen transects were chosen across buffer and transition zones of Indian Sundarban, each equipped with 4–6 sedimentation pins, to gauge accretion rates in premonsoon, monsoon, and postmonsoon season during period March 2020 to March 2022, considering diverse hydro-geomorphic attributes. The vertical accretion rate exhibits a curvilinear pattern, ascending from the lower seafront zone towards the middle of the estuarine region, reaching a peak, and subsequently declining towards the upper estuarine zones. Specifically, in the seafront zone, the yearly mean accretion (focusing solely on accretion and excluding subsidence) ranges between 5 and 10 mm/year, while in the middle estuarine region, it increases to an average of 38.1 ± 9.44 mm/year. However, in the upper zone, the yearly mean accretion decreases slightly to approximately 17.71 ± 9.54 mm/year. Similar patterns are observed when considering the distance from the riverfront to the interior of the mangrove zones. Seasonal analysis demonstrates that the postmonsoon period exhibits the highest accretion, followed by the premonsoon season, while the monsoon tends to be erosional in nature. Mangrove zonation, and subsequent distribution of root height, root diameter, root density, tree density, tree basal area, inundation frequency, current speed, variation in suspended sediment concentration were identified as influential factors affecting vertical accretion rates. Based on multiple regression analysis, root density, inundation frequency, and tree density demonstrated moderate-positive correlations with vertical accretion, whereas distance from the riverfront, root height, and root diameter displayed negative correlations. This study enhances our understanding of the intricate processes behind vertical accretion in the Sundarban, offering valuable insights for future research and advancing discussions on the broader implications of climate change in coastal regions.

A Privacy-Preserving Data Mining Through Comprehensive GNIPP Approach in Sensitive Data Sets

The quick growth of methods for analyzing data and the availability of easily available datasets have made it possible to build a thorough analytics model that can help with support decision-making. In the meantime, protecting personal privacy is crucial. A popular technique for medical evaluation and prediction, decision trees are easy to comprehend and interpret. However, the decision tree construction procedure may reveal personal information about an individual. By keeping the statistical properties intact and limiting the chance of privacy leaking within a reasonable bound, differential privacy offers a formal mathematical definition of privacy. To construct a boosting random forest that preserves privacy, we propose a Gaussian Noise Integrated Privacy Preservation (GNIPP) in this study. To address the issue of personal information breaches, we have designed a unique Gaussian distribution mechanism in GNIPP that enables the nodes with deeper depth to obtain more privacy during the decision tree construction process. We propose a comprehensive boosting technique based on the decision forest's prediction accuracy for assembling multiple decision trees into a forest. Furthermore, we propose an iterative technique to accelerate the assembly of decision trees. After all, we demonstrate through experimentation that the suggested GNIPP outperforms alternative algorithms on two real-world datasets.

Regression of Concurrence via Local Unitary Invariants

Concurrence is a crucial entanglement measure in quantum theory used to describe the degree of entanglement between two or more qubits. Local unitary (LU) invariants can be employed to describe the relevant properties of quantum states. Compared to quantum state tomography, observing LU invariants can save substantial physical resources and reduce errors associated with tomography. In this paper, we use LU invariants as explanatory variables and employ methods such as multiple regression, tree models, and BP neural network models to fit the concurrence of 2-qubit quantum states. For pure states and Werner states, by analyzing the correlation between data, a functional formula for concurrence in terms of LU invariants is obtained. Additionally, for any two-qubit quantum states, the prediction accuracy for concurrence reaches 98.5%.

Fruit and Vegetable Image Recognition Based on Multiple Tree Models: Applications of Random Forest, XGBoost and Decision Tree

The primary objective of this study is to evaluate and compare the performance of three machine learning models— Random Forest, XGBoost, and Decision Tree—in the context of fruit and vegetable image classification. This research aims to identify which model best handles the challenges associated with imbalanced datasets and complex data structures. The ultimate goal is to contribute to the development of more efficient and accurate automated systems for agricultural applications, thereby improving productivity and reducing operational costs in the industry. This study utilized a dataset of 3,825 images covering 36 fruit and vegetable classes. Images were resized, normalized, and augmented to enhance diversity. Three models—Random Forest, XGBoost, and Decision Tree—were trained on this dataset. Performance was evaluated using accuracy, precision, recall, and F1-score to assess classification effectiveness and handling of class imbalances. The evaluation revealed that XGBoost outperformed Random Forest and Decision Tree in fruit and vegetable image classification, achieving the highest accuracy of 96.66%. XGBoost demonstrated superior handling of class imbalances and complex data structures, reflected in its precision and recall scores across various classes. Random Forest also performed well, closely following XGBoost, while Decision Tree exhibited more variability in results, indicating potential overfitting in certain classes. In conclusion, this study highlights the effectiveness of ensemble methods, particularly XGBoost, in agricultural image classification tasks. These findings suggest that XGBoost is a robust model for similar applications, offering improved accuracy and reliability

Application of machine learning for predictions of consecutive dependent data of type {[(a, b)->c]->d}

Objective: Machine learning techniques have emerged in response to the desire for automatic pattern detection withindatasets in fields such as statistics, mathematics, and data analytics. They allow for the extraction of relevant informationfrom datasets of significantly large volumes, providing the possibility of making predictions. This paper presents an application focused on decision trees, linear regression, and random forest regression algorithms to predict final data fromconsecutive dependent data of type {[(a, b) → c] → D}. Methodology: The study adopts a quantitative research design, which takes as input datasets based on interval data. It utilizes a correlational research model by implementing Python and its Scikit-Learn library, which includes various algorithms for prediction. Specifically, we compare the application of decision trees, linear regression, and random forest regression on the same set of datasets, but with a characteristic of dependency between them. Results: Upon application of the proposed model, it yields an estimated prediction score, which indicates the accuracy of the model concerning the data provided. Conclusions: The application of a complex algorithm does not inherently guarantee a higher rate of accuracy. Conversely, configuring the model correctly, training multiple trees, or adjusting parameter values can significantly enhance the obtained results

A Hybrid Machine Learning‐Based Data‐Centric Cybersecurity Detection in the 5G‐Enabled IoT

ABSTRACTBy increasing smart objects as high‐potential computing devices and 5G technology, the Internet of Things (IoT) has emerging technology to provide a data‐centric infrastructure for detecting safe device‐to‐device communications by supporting security and privacy issues. As a 5G‐enabled communication technology, vehicle‐to‐vehicle (V2V) communication provides a wireless exchange information connection between smart vehicles, intelligent devices, and a cloud‐edge computing environment. This connection should be established with a safe and secured run‐time protection system to avoid many critical anomalies and misbehavior problems. Detecting run‐time malicious transformations with data‐centric misbehaving reactions is a main challenge for autonomous vehicle communications with 5G‐enabled communication technology. This paper provides a hybrid genetic algorithm‐based ensemble bagged trees (GA‐EBT) algorithm for a data‐centric misbehavior detection approach to support the V2V communications against malicious and misbehavior transactions. In this hybrid algorithm, GA provides a beneficial performance on the training accuracy factor based on a single‐objective fitness function and EBT supports a high‐degree prediction and training process using multiple decision tree models with bootstrapped samples of the training data. For the evaluation of the proposed algorithm, four real test cases are applied for messaging injection attacks in the V2V environments in comparison to the state‐of‐the‐art machine learning algorithms. The experimental results show that the proposed hybrid approach can achieve an optimal high rate accuracy factor of 99.999, precision and recall factors of 100%, and an F1‐Score factor of 100% to detect unexpected cyber‐attacks for the V2V communications in the IoT environment.

Heart disease prediction using machine learning models

Abstract. Heart disease remains a global health threat, making rapid identification crucial. Using the Heart Disease Dataset from Kaggle, this research employs a Random Forest model to analyze 13 clinical variables from 1,025 samples. To improve accuracy and address class imbalance, the dataset was split into training and test sets, utilizing methods such as Z-scores, SMOTE, and feature selection. The Random Forest model, which combines multiple decision trees, achieved high performance with an accuracy of 98.54%, identifying key predictors such as chest pain type, maximum heart rate, and thalassemia. Compared to a single decision tree, the Random Forest model reduces overfitting, improves generalization, and increases predictive accuracy. Factors like cholesterol levels, resting blood pressure, and exercise-induced angina were also considered. By averaging results from multiple trees, the model offers reliable and stable predictions, highlighting its potential in clinical settings for early detection and personalized treatment strategies. This study aims to assist healthcare providers in better allocating resources, planning preventive measures, and tailoring treatment plans to individual patients.

On-Tree Mango Fruit Count Using Live Video- Split Image Dataset to Predict Better Yield at Pre-Harvesting Stage

This study introduces a method for fruit counting in agricultural settings using video capture and the YOLOv7 object detection model. By splitting captured videos into frames and strategically selecting representative frames, the approach aims to accurately estimate fruit counts while minimizing the risk of double counting. YOLOv7, known for its efficiency and accuracy in object detection, is employed to analyze selected frames and detect fruits on trees. Demonstrated the method's effectiveness through its ability to provide farmers with precise yield estimations, optimize resource management, and facilitate early detection of orchard issues such as pest infestations or nutrient deficiencies. This technological integration reduces labor costs and supports sustainable agricultural practices by improving productivity and decision-making capabilities. The scalability of the approach makes it suitable for diverse orchard sizes and types, offering a promising tool for enhancing agricultural efficiency and profitability. The researcher compared YOLOv5n, YOLOv5s, YOLOv7, and YOLOv7-tiny with eight-sided imaging techniques around the tree. The experimental results of YOLOv7 with the eight-sided technique performed best and achieved a count accuracy of 97.7% on a single tree in just 17.112 ms of average inference time. On multiple trees, it is 95.48% in just 17 ms of average inference time, with the help of an eight- sided method on tree images.

A classification and recognition model for multiple fruit tree leaf diseases

Abstract Fruit tree leaf diseases affect both fruit survival rate and orchard revenue. Rapid and accurate identification of tree leaf diseases not only prevents the spread of diseases but also ensures healthy tree growth and improves fruit quality. Existing models mostly focus on individual types of tree leaf diseases, with limited applications in the identification of multiple tree leaf diseases. In response to the insensitivity of current classification models to disease region features and the issue of increased error rates due to the presence of similar diseases, this study proposes a residual network model based on the ECA channel attention mechanism and meta-Acon adaptive activation function.The model employs ResNet34 as the backbone network and incorporates the ECA module after each residual block to focus on relevant information. Additionally, a new activation function called meta-Acon is introduced to enhance the model's generalization ability through its dynamic learning capability. Finally, the model's recognition performance is improved by fusing bottom-level features with features from other layers using the Feature Pyramid Network (FPN).Experimental results on a dataset augmented with Mosaic processing show that the FPEM-ResNet34 model achieves a classification accuracy of 98.46%. Compared to other common models such as VGG-16, Inception-V1,AlexNet, and ResNet50, the proposed method mentioned in this paper demonstrates more effective improvement in the accuracy of tree leaf classification, making it highly valuable for practical applications..

Estimation Stand Volume, Basal Area and Quadratic Mean Diameter Using Landsat 8 OLI and Sentinel‐2 Satellite Image With Different Machine Learning Techniques

ABSTRACTThe data required for sustainable forest planning is provided by traditional forest inventories, which are labor, time, and cost‐intensive. Providing this data quickly, reliably, and accurately is crucial for planners and researchers. The objective of this study was to predict stand basal area (BA), stand volume (V), and quadratic mean diameter (dq) by leveraging vegetation indices (VIs) and reflectance (R) derived from Landsat 8 OLI and Sentinel 2 satellite images, along with topographic (T) data obtained from ALOS‐PALSAR satellite imagery. Forest inventory data for a total of 250 sample plots were used for modeling in the study. Stand parameters were estimated using support vector machines (SVM), multiple linear regression (MLR), decision tree (DT), and random forest (RF) algorithms. In modeling V, BA, and dq, both individual and combinations of R, VIs, and T values obtained from satellite imagery were used as independent variables. Using the generated datasets, each of the stand parameters was modeled separately with MLR, SVM, RF, and DT algorithms, and the success of the models was compared to determine the modeling technique and dataset with the highest success for the relevant parameter. The results showed that for each stand parameter, the highest model success was achieved in the combined dataset, which was created by combining all datasets. However, in terms of modeling techniques, the highest success for each stand parameter was achieved with different modeling techniques. The highest success for V is obtained in the model using the SVM method (R2 = 0.78; RMSE = 0.28 m3/ha), the RF method yielded the highest model performance for BA (R2 = 0.70; RMSE = 2.53 m2/ha), and finally, the highest success for dq was obtained in the DT method (R2 = 0.74; RMSE = 0.02 cm). In general, the datasets obtained from Sentinel 2 images showed higher model success than the datasets obtained from Landsat 8 OLI images.

RS-Net: Hyperspectral Image Land Cover Classification Based on Spectral Imager Combined with Random Forest Algorithm

Recursive neural networks and transformers have recently become dominant in hyperspectral (HS) image classification due to their ability to capture long-range dependencies in spectral sequences. Despite the success of these sequential architectures, mainstream deep learning methods primarily handle two-dimensional structured data. However, challenges such as the curse of dimensionality, spectral variability, and confounding factors in hyperspectral remote sensing images limit their effectiveness, especially in remote sensing applications. To address this issue, this paper proposes a novel land cover classification algorithm that integrates random forests with a spectral transformer network structure (RS-Net). Firstly, this paper presents a combination of the Gramian Angular Field (GASF) and Gramian Angular Difference Field (GADF) algorithms, which effectively maps the multidimensional time series constructed for each pixel onto two-dimensional image features, enabling precise extraction and recognition in the backend network algorithms and improving the classification accuracy of land cover types. Secondly, to capture the relationships between features at different scales, this paper proposes a SpectralFormer network architecture using the Context and Structure Encoding (CASE) module to effectively learn dependencies between channels. This architecture enhances important features and suppresses unimportant ones, thereby addressing the semantic gap and improving the recognition capability of land cover features. Finally, the final prediction results are determined by a voting mechanism from the Random Forest algorithm, which synthesizes predictions from multiple decision trees to enhance classification stability and accuracy. To better compare the performance of RS-Net, this paper conducted extensive experiments on three benchmark HS datasets obtained from satellite and airborne imagers, comparing various classic neural network models. Surprisingly, the RS-Net algorithm achieves high performance and efficiency, offering a new and effective tool for land cover classification.

Study on Audit Risk Model Based on Data Mining Algorithm

With the widespread application of modern information technologies such as big data, enterprises have generated a large amount of data in their daily operations. The traditional audit methods have drawbacks such as high cost and high consumption, and can no longer meet the needs of audit work in the digital era. Therefore, it is urgent to adopt audit methods suitable for the digital era to improve audit quality and reduce enterprise Audit risk. For professional auditors in enterprises, it is very important to utilize emerging technologies such as data mining algorithms. The audited enterprise may tamper with its financial statements, and identifying high-quality audit data from massive amounts of data is a huge challenge. Compared to traditional audit methods, data mining algorithms can significantly improve the efficiency and accuracy of audit work. For example, the BP neural network, with its powerful nonlinear mapping ability, can capture complex relationships in data; Support vector machines classify data by finding the optimal hyperplane in high-dimensional space and have good generalization ability; random forest reduces overfitting and improves prediction accuracy by integrating multiple decision trees; Association rule algorithms can discover interesting relationships between data items, helping auditors identify potential fraudulent behavior or abnormal transactions. Therefore, the purpose of this study is to accurately evaluate and reduce the audit risk of enterprises, and to build an audit risk model using computer data mining algorithms. This provides necessary reference and guidance for auditors to conduct data analysis and mine valuable data during the audit process of enterprises, thereby improving audit efficiency.

Phylogenomics reveal species limits and inter-relationships in the narrow-range endemic lycian salamanders

Salamanders of the genus Lyciasalamandra are represented by as many as 20 narrow-range endemic taxa inhabiting the Mediterranean coast of Turkey and a handful of Aegean Islands. Despite recent molecular phylogenetic studies, the genus is rife with uncertainty about the number of contained species and their phylogenetic relationships, both of which can interfere with needed conservation actions. To test species limits and infer interrelationships we generated as many as 113,176 RAD loci containing 229,427 single nucleotide polymorphisms (SNPs), for 110 specimens of Lyciasalamandra representing 19 of the 20 described taxa. Through a conservative species delimitation approach, we found support for eight species in the genus which broadly agree with currently described species-level diversity. We then use multiple coalescent-based species tree methods to resolve relationships in this relatively old, synchronous species radiation. We recommend synonymization of the largely over-split subspecific taxa, and the elevation of L. luschani finikensis to full species status as L. finikensis. Our hope is that this revised taxonomic framework provides a stable foundation for conservation management in these fragile, microendemic taxa.

Fault traceability of power grid dispatching system based on DPHS-MDS and LambdaMART algorithm

Under the background of increasing system service data, it is difficult to trace the faults of power dispatching system. The current fault tracing method has some problems, such as low precision, low efficiency and difficult troubleshooting. To solve this problem, a fault tracing method based on data partition hybrid sampling method and multiple incremental regression tree algorithm is proposed. In this paper, it first uses the hybrid sampling method of data partition and dynamic selection technology to detect the business anomaly, and then applies the clustering algorithm and the information difference graph model to realize the fault tracing of system components. The experimental results showed that the F-metric value and geometric mean value of the study design method were 0.964 and 0.685, respectively. In addition, normalized discounted cumulative gains were observed in the top 10, and the mean average precision of the top 10 was 0.752 and 0.186, respectively. The proposed method can effectively improve the fault tracing efficiency of power grid operation and maintenance personnel, and provide strong data support for the safety maintenance of power grid dispatching system.

Development of a unified framework of low-rank approximation and deep neural networks for predicting the spatial variability of SSC in `Spania' watermelons using vis/NIR hyperspectral imaging

Soluble Solids Content (SSC) is an important quality attribute that represents the internal quality of fruits. Visible and near-infrared (Vis/NIR) combined with chemometric algorithms are now popular methods for non-invasive measurement and visualization of SSC. However, in fruits with a thick rind and a large flesh volume, such as watermelon, SSC is not evenly distributed across the fruit. The variability in SSC across the fruit flesh may have an adverse effect on the accuracy of quality analysis algorithms. Thus, this paper presents an accurate and efficient approach for predicting the spatial variation of SSC in watermelons using a combined framework of low-rank approximation and deep neural networks. Watermelon ‘Spania’ was selected for this study, and hyperspectral images of each watermelon were taken from various views, including the top, bottom, and two lateral views. The low-rank property is employed as a constraint in this approach to eliminate the unwanted variations in the spectral data. The low-rank component of the spectral data, free of unwanted variations, is then fed into a fully connected neural network (FNN) for the prediction of watermelon SSC values. The proposed approach obtained optimal performance in calibration and prediction with RC2=0.982, RMSEP = 0.132, and RP2=0.945, RMSEP = 0.195 respectively. Further, the prediction outcomes of the proposed method were compared with state-of-the-art such as Partial Least Square Regression (PLSR), Support Vector Regression (SVR), Multiple Linear Regression (MLR), Decision Tree (DT), and Random Forest (RF). The overall results showed that combining HSI data with a low-rank and deep neural network framework is an efficient method for accurately predicting SSC in watermelons as well as correctly predicting spatial variability of SSC in individual fruits.

Corruption prevention in organizational clustering in Indonesia: through the role of the HU-model in detecting corruption

PurposeThis paper aims to test the effectiveness of the Haryono Umar (HU)-model used in corruption prevention strategies through corruption detection as a tool for detecting corruption because the mode of corruption is increasingly dynamic and complex by focusing on the causes of corruption: pressure, opportunity, rationalization, capability and lack of integrity.Design/methodology/approachThe research uses multiple regression methods, classification and regression trees and the HU-model application system developed by researchers. The research sample uses secondary data from financial reports on the Indonesia stock exchange according to organizational clustering (such as red, grey and green areas).FindingsThe research result showed that of the 470 sample companies, there were 445 companies, or 98.9%, in the red cluster (indicated corruption), 19 companies, or 4.04, in the green clusters or not indicated corruption and six companies, or 1.28%, were included in the grey cluster or potential corruption. By knowing the cluster of an organization, efforts to prevent corruption can be made effective and efficient. Implementing the HU-model proves that the amount of pressure, the abundance of opportunities, the ease of rationalization and the high level of position and authority strengthen the drive for corruption if there is a lack of integrity.Research limitations/implicationsEach internal organization can use this model independently and find conditions related to corruption so that they can immediately take action to prevent it.Originality/valueThe application of the HU-model is a discovery in preventing corruption by focusing on the possibility of corruption occurring in each organization through organizational clustering.

A Methodology for Deciding on Well Seal Options for Abandonment

The energy transition to achieve net zero anthropogenic CO 2 emissions will require permanently sealing and abandoning wells or boreholes drilled for many different purposes, including: underground CO 2 storage; hydrogen storage; geothermal energy extraction; site characterisation for radioactive waste repositories; and hydrocarbon extraction. It is necessary to objectively decide what sealing materials, combination of materials and methods of emplacement are optimal for a given well sealing project, taking into account the properties of lithologies penetrated, the physical and geochemical conditions around the well, the geometry of the well, and characteristics of well engineering. A workflow that would allow for a structured approach to designing well seals during well decommissioning and a set of complementary tools that will enable an informed, structured, transparent and traceable process for designing a well decommissioning sealing solution for a particular well in a given geological environment is presented in this paper. The tools include: 1) a database of Features Events and Processes; 2) a decision tree for evaluating different well sealing options and for documenting the rationale for decisions; 3) an approach for comparing options using multiple decision trees; 4) and a suite of simplified numerical models to inform judgements.

Research on the influencing factors of housing price based on multiple linear regression and random forest

This paper aims to use multiple linear regression model and random forest models to analyze and study the factors affecting the housing price in Boston. The multiple linear regression model describes the relationship between multiple independent variables and one dependent variable through linear equations, and the random forest improves the accuracy and robustness by constructing multiple decision trees and combining their prediction results. To deal with complex nonlinear relationships and high dimensional data. Housing price is an important index to reflect the level and condition of economic and social development of a region, so it is of theoretical value and practical significance to explore its influencing factors and ways and degrees. Multiple factors are selected to analyze the weight and importance of each influencing factor, so as to help the government and decision makers to formulate more accurate policies, promote the stable development of the market, and provide scientific decision-making support for real estate developers, investors and ordinary buyers. In this study, the random forest model based on decision tree was used to clean, select and reduce the acquired housing price data, and to find out the main factors affecting housing price from the perspective of information gain, so as to obtain a relatively complete mathematical model and provide a reference scheme for future research by scholars.

Predicting the Views of TED Talks using Machine Learning Model

A forecast of such accuracy would have significant implications in this respect for content creators, marketers, and researchers alike. The paper questions whether machine learning models can predict the views that any certain TED Talks receive with a dataset containing features on the duration of the talks, characteristics of the speaker, the date of publication, metrics indicating audience engagement, and so on. In this work, we employ techniques of multiple regression, decision trees, and ensemble methods to identify empirically which factors have the most influence and to come up with the best models for the prediction. Indeed, the results obtained easily indicate that the ensemble models-especially Random Forest and Gradient Boosting- perform way above the rest. These features help to emphasize the fact that machine learning will help guide a content strategy to better reach audiences and give insight into a deeper understanding of how digital content is consumed. Keywords: TED Talks, machine learning, predictive modeling, view prediction, Random Forest, Gradient Boosting, digital content analysis, audience engagement.