Random Forest Regression Algorithm Research Articles

Reconstruction of Hourly FY-4A AGRI Land Surface Temperature under Cloud-Covered Conditions Using a Hybrid Method Combining Spatial and Temporal Information

Land Surface Temperature (LST) products obtained by thermal infrared (TIR) remote sensing contain considerable blank areas due to the frequent occurrence of cloud coverage. The studies on the all-time reconstruction of the cloud-covered LST of geostationary meteorological satellite LST products are relatively few. To accurately fill the blank area, a hybrid method for reconstructing hourly FY-4A AGRI LST under cloud-covered conditions was proposed using a random forest (RF) regression algorithm and Savitzky-Golay (S-G) filtering. The ERA5-Land surface cumulative net radiation flux (SNR) reanalysis data was first introduced to represent the change in surface energy arising from cloud coverage. The RF regression method was used to estimate the LST correlation model based on clear-sky LST and the corresponding predictor variables, including the normalized difference vegetation index (NDVI), the normalized difference water index (NDWI), surface elevation and slope. The fitted model was then applied to reconstruct the cloud-covered LST. The S–G filtering method was used to smooth the outliers of reconstructed LST in the temporal dimension. The accuracy evaluation was performed using the measured LST of the representative meteorological stations after scale correction. The coefficients of determination derived with the reference LST were all above 0.73 on the three examined days, with a bias of −1.13–0.39 K, mean absolute errors (MAE) of 1.46–2.4 K, and root mean square errors (RMSE) of 1.77–3.2 K. These results indicate that the proposed method has strong potential for accurately restoring the spatial and temporal continuity of LST and can provide a solution for the production and research of gap-free LST products with high temporal resolution.

Optimization of process parameters of selective laser melted nickel-based superalloy for densification by random forest regression algorithm and response surface methodology

In this study, Random Forest Regression (RFR) and Response Surface Methodology (RSM) were employed to predict the optimized processing parameters to achieve the highest densification of a nickel-based superalloy Mar-M247LC fabricated by selective laser melting (SLM). The RFR model considered input processing parameters such as laser power, hatch distance, and scanning speed. A dataset of 223 samples, was used to train the RFR model. As a result, the RFR model exhibited accuracy of 99.57 %, R2 value of 0.976, Mean Square Error (MSE) of 0.402, and Mean Absolute Percentage Error (MAPE) of 0.426 % on testing set. In addition to the RFR model, this study also employed Central Composite Design (CCD) and RSM to optimize the processing parameter sets. Subsequently, this study conducted Box-Behnken Design (BBD) to experimentally validate the accuracy of this RFR model. In the end, a set of the optimal processing parameters was tested and resulted a sample densification of 99.959 %, outperformed that in the original database before building the RFR model, which was 99.734 %. In summary, the RFR models was able to predict densification with accuracy, and by coupling with RSM, the optimal processing parameter could be obtained, so better densification of the build could be achieved.

Optical frequency multiplication using residual network with random forest regression

In this work, we present a method for optical frequency multiplication utilizing a hybrid deep learning approach that integrates the Residual Network (ResNet) with the Random Forest Regression (RFR) algorithm. Three different frequency multiplication modulation schemes are adopted to illustrate the method, which can obtain suitable parameters for these schemes. Based on the parameters predicted by the algorithm, the 8-tupling, 12-tupling, and 16-tupling mm-wave signals are generated by numerical simulation. The simulation results show that for 8-tupling frequency multiplication, an OSSR (optical sideband suppression ratio) is 30.73 dB and an RFSSR (radio frequency spurious suppression ratio) of 80 GHz is 42.29 dB. For 12-tupling frequency multiplication, the OSSR is 30.09 dB, and the RFSSR of the 120 GHz mm wave is 36.21 dB. For generating 16-tupling frequency mm-wave, an OSSR of 29.86 dB and an RFSSR of 34.52 dB are obtained. In addition, the impact of amplitude fluctuation and bias voltage drift on the quality of mm-wave signals is also studied.

Breast Cancer Disease Prediction Using Random Forest Regression and Gradient Boosting Regression

The current research endeavors to evaluate the efficacy of regression-based machine learning algorithms through an assessment of their performance using diverse metrics. The focus of our study involves the implementation of the breast cancer Wisconsin (Diagnostic) dataset, employing both the random forest and gradient-boosting regression algorithms. In our comprehensive performance analysis, we utilized key metrics such as Mean Squared Error (MSE), R-squared, Mean Absolute Error (MAE), and Coefficient of Determination (COD), supplemented by additional metrics. The evaluation aimed to gauge the algorithms' accuracy and predictive capabilities. Notably, for continuous target variables, the gradient-boosting regression model emerged as particularly noteworthy in terms of performance when compared to other models. The gradient-boosting regression model exhibited remarkable results, highlighting its superiority in handling the breast cancer dataset. The model achieved an impressively low MSE value of 0.05, indicating minimal prediction errors. Furthermore, the R-squared value of 0.89 highlighted the model's ability to explain the variance in the data, affirming its robust predictive power. The Mean Absolute Error (MAE) of 0.14 reinforced the model's accuracy in predicting continuous outcomes. Beyond these core metrics, the study incorporated additional measures to provide a comprehensive understanding of the algorithms' performance. The findings underscore the potential of gradient-boosting regression in enhancing predictive accuracy for datasets with continuous target variables, particularly evident in the context of breast cancer diagnosis. This research contributes valuable insights to the ongoing exploration of machine learning algorithms, providing a basis for informed decision-making in medical and predictive analytics domains.

A MACHINE LEARNING BASED SYSTEM TO CLASSIFY FOREST FIRE PREDICTION FOR FOREST SAFETY

Forecasting forest fires requires the application of different techniques and instruments to evaluate the likelihood and potential seriousness of a fire starting in a forested region. Elements like drought, extreme temperatures, and human behaviors like bonfires, cigarettes, and pyrotechnics can all play a role in sparking forest fires. Various methods such as statistical examination, AI algorithms, and satellite pictures are employed to gather and dissect data on climate conditions, moisture levels, topography, and other factors that may heighten the chance of a forest fire. Models for predicting forest fires can furnish advanced caution systems to inform officials and locals of potential fire hazards. Predicting forest fires in the future is expected to decrease the impact of fires. This study focuses on developing a system for predicting forest fires, which calculates the likelihood of a fire starting based on various meteorological factors such as location (latitude and longitude) and temperature. The Random Forest regression algorithm was utilized to create this predictive model. Key Words: Forest Fire, Machine Learning, Temparature, Random Forest, Regression

Prediction improvement of compressive strength and strain of directionally solidified TiAl alloy based on training data size adjustment

Compressive strength and compressive strain, which are important mechanical properties of directionally solidified TiAl alloy, were predicted using machine learning algorithms, specifically Multiple Linear Regression (MLR), and Random Forest Regression (RFR). The input variables for the machine learning model were designated as the composition of the directionally solidified TiAl alloy, experimental parameters (pulling velocity, input power), and compression test parameters (strain rate, compression temperature). Compressive strength and compressive strain were designated as the output variables. Although the typical ratio of training and test data is 8:2, this study used different ratios of 9:1, 7:3, and 6:4 for machine learning, and excellent R2 values were obtained for all ratios. The feature importance, which can identify the factor that has the most influence on the output variables, was obtained through the RFR algorithm. According to the feature importance, temperature was found to have the greatest influence on compressive strength, while the Erbium (Er) element had the most significant influence on compressive strain. Through the results of feature importance, it was possible to quantitatively investigate the relationship between the Er element, a microalloying element that affects the microstructure of TiAl alloy, and the compressive properties. Furthermore, the study was conducted on which data ratio between training and test data is most suitable for predicting the compressive strength and strain of TiAl alloy.

Powerlifting total score prediction based on an improved random forest regression algorithm

This paper discusses the use of an improved random forest regression algorithm (RFRA) to predict the total score of powerlifters. The paper collected the age, weight, and total score of multiple powerlifters, and then used an improved RFRA to build a predictive model. The parameters of this model are optimized by a differential squirrel search algorithm. The experimental results show that our proposed method can effectively predict the total score of powerlifters with an error of less than 10%, which can provide a reference for experts and athletes before training or competition.

A high-resolution map of diffuse groundwater recharge rates for Australia

Abstract. Estimating groundwater recharge rates is important to understand and manage groundwater. Numerous studies have used collated recharge datasets to understand and project regional- or global-scale groundwater recharge rates. However, recharge estimation methods all have distinct assumptions, quantify different recharge components and operate over different temporal scales. We use over 200 000 groundwater chloride measurements to estimate groundwater recharge rates using an improved chloride mass balance (CMB) method across Australia. Groundwater recharge rates were produced stochastically using gridded chloride deposition, runoff and precipitation datasets. After filtering out groundwater recharge rates where the assumptions of the method may have been compromised, 98 568 estimates of recharge were produced. The resulting groundwater recharge rates and 17 spatial datasets were integrated into a random forest regression algorithm, generating a high-resolution (0.05°) model of groundwater recharge rates across Australia. The regression reveals that climate-related variables, including precipitation, rainfall seasonality and potential evapotranspiration, exert the most significant influence on groundwater recharge rates, with vegetation (the normalised difference vegetation index or NDVI) also contributing significantly. Importantly, the mean values of both the recharge point dataset (43.5 mm yr−1) and the spatial recharge model (22.7 mm yr−1) are notably lower than those reported in previous studies, underscoring the prolonged timescale of the CMB method, the potential disparities arising from distinct recharge estimation methodologies and limited averaging across climate zones. This study presents a robust and automated approach to estimate recharge using the CMB method, offering a unified model based on a single estimation method. The resulting datasets, the Python script for recharge rate calculation and the spatial recharge models collectively provide valuable insights for water resource management across the Australian continent, and similar approaches can be applied globally.

Utilizing the Fusion Characteristics of Multispectral and Electronic Noses to Detect Soil Main Nutrient Content

Soil nutrients are essential for plant growth, and it is crucial to accurately detect their levels. However, current methods for detecting soil nutrients still have low accuracy and high costs. In this study, we propose a multi-source fusion feature approach that combines multispectral image technology with electronic nose gas response technology to achieve rapid, high-precision, and cost-effective detection of soil nutrients, including soil organic matter (SOM), total nitrogen (TN), available potassium (AK), and available phosphorus (AP). To begin, a multispectral camera was used to collect spectral images and extract seven index features. Additionally, an electronic nose was used to collect soil pyrolysis gases to obtain the response curve and extract seven response features. These two sets of features were effectively fused to form a 106 × 98 fusion feature space. Secondly, principal component analysis (PCA) and Pearson correlation coefficient (PCC) algorithms were used to optimize and reduce the dimensionality of the fused feature space. Finally, a predictive model of the relationship between the feature space and the nutrient content of the soil samples was established using the random forest (RF) and partial least squares regression (PLSR) algorithms. The predictive performance of the model was used to evaluate the accuracy of soil nutrient detection. The results showed that the PLSR modeling of the optimized feature space of PCA achieved high accuracy in predicting the levels of main soil nutrients, with R2 values for SOM, TN, AK, and AP of 0.96, 0.95, 0.84, and 0.73, and the RMSE values of 0.56, 0.07, 3.21, and 3.70, respectively. Compared to using only electronic nose gas response technology, the soil nutrient detection with multi-source data features method in this study not only improved the accuracy for SOM and TN but also for metal elements AK and AP in soil. The spectral index features proposed in this study were able to compensate for the limitations of the electronic nose response features, and modeling with the fusion feature space resulted in an accurate prediction of SOM, TN, AK, and AP levels.

Tracking and tracing water consumption for informed water sensitive intervention through machine learning approach

To develop a water conscious strategy, it is critical to track and trace water from its source to the end users, understand water conservation behaviors, and identify the factors that influence water consumption. However, in developing nations, little research has been done to provide a quantitative picture of how water is consumed and transformed in urban households, as well as the water sensitive interventions needed to improve access to clean water. Hence, the main objective of the study was to determine the most significant residential water consumption variables and to predict residential water consumption in a way that can generate water consumption information for water sensitive intervention decision making using the case study of Adama city in Ethiopia. A combination of top down and bottom up data collection techniques were employed as the data collection instrument. Machine learning was integrated with spatial and socioeconomic analytic techniques to estimate daily household water consumption and identify the factors that significantly influence household water consumption. The results show that there is only “one source option” for the city’s clean water supply and that different water harvesting methods are not likely to be developed. The average daily water consumption per person is 69 liters which falls below the national standard of 80 liters allocated per person per day. The result reveals that the water distribution network covers only 45% of the city master plan. About 38% of the water demand is unmet and 30% of households only receive water once every three days or fewer. This shows that the city is experiencing physical and economic water scarcity. The results demonstrated that family size, housing quality, income, number of rooms, legal status of the parcel, supply reliability, climate, and topographical features are the most important factors in predicting residential water consumption. This study further demonstrates how well supervised machine learning models, such as the Random Forest Regression algorithm, can predict the household’s daily water consumption. The findings also showed that there is a need for significant improvements in water saving habits of the households. Another conclusion that can be drawn is that as long as the city’s business as usual water consumption practice doesn’t change, the water supply problem will worsen over time.

Machine Learning Prediction of Quantum Yields and Wavelengths of Aggregation-Induced Emission Molecules.

The aggregation-induced emission (AIE) effect exhibits a significant influence on the development of luminescent materials and has made remarkable progress over the past decades. The advancement of high-performance AIE materials requires fast and accurate predictions of their photophysical properties, which is impeded by the inherent limitations of quantum chemical calculations. In this work, we present an accurate machine learning approach for the fast predictions of quantum yields and wavelengths to screen out AIE molecules. A database of about 563 organic luminescent molecules with quantum yields and wavelengths in the monomeric/aggregated states was established. Individual/combined molecular fingerprints were selected and compared elaborately to attain appropriate molecular descriptors. Different machine learning algorithms combined with favorable molecular fingerprints were further screened to achieve more accurate prediction models. The simulation results indicate that combined molecular fingerprints yield more accurate predictions in the aggregated states, and random forest and gradient boosting regression algorithms show the best predictions in quantum yields and wavelengths, respectively. Given the successful applications of machine learning in quantum yields and wavelengths, it is reasonable to anticipate that machine learning can serve as a complementary strategy to traditional experimental/theoretical methods in the investigation of aggregation-induced luminescent molecules to facilitate the discovery of luminescent materials.

Recurrent Neural Networks and classical machine learning methods for concentrations prediction of aluminum alloy in laser Induced breakdown spectroscopy

Recurrent Neural Networks are classes of Artificial Neural Networks that establish connections between various nodes in a directed or undirected graph for investigation of the temporal dynamical. In this study, different Recurrent Neural Network (RNN) architectures are utilized for quantitative analysis of aluminum alloys by the laser induced breakdown spectroscopy (LIBS) technique. The fundamental harmonic (1064 nm) of a nanosecond Nd:YAG laser pulse is employed to generate the LIBS plasma for the prediction of constituent concentrations of the aluminum standard samples. For the purpose of predicting concentration, Recurrent Neural Networks based on different networks, such as Long Short Term Memory (LSTM), Gated Recurrent Unit (GRU), Simple Recurrent Neural Network (Simple RNN), and as well as Recurrent Convolutional Networks comprising of Conv-SimpleRNN, Conv-LSTM and Conv-GRU are employed. Then, a comparison is made among prediction by classical machine learning methods of support vector regressor (SVR), the Multilayer Perceptron (MLP), Decision Tree algorithm, Gradient Boosting Regression (GBR), Random Forest Regression (RFR), Linear Regression, and k-Nearest Neighbor (KNN) algorithm. Results demonstrated that the machine learning tools based on Convolutional Recurrent Networks had the best efficiencies in prediction of the most of the elements among other multivariate methods.

Machine-learning-assisted hydrogen adsorption descriptor design for bilayer MXenes

Currently, most of the MXene hydrogen storage materials with excellent performances are screened by empirical trial-and-error methods. All of them are single-layer materials, and they have difficulty meeting actual demands. Herein, we report the accurate prediction of hydrogen adsorption energies for three adsorption modes inside M12X1–M22X2 bilayer MXenes using only physical intrinsic features (no density functional theory computational variables). The gradient boosting regression and random forest regression algorithms achieved R2 of 0.957/0.946 and 0.952/0.935 for chemisorption and physical adsorption models on the training/test set, respectively. In particular, the presence of a nanopump effect mechanism in the MXenes with a small layer spacing ensured that the system had a strong Kubas adsorption of H2. Symbolic regression was used to guide the design of hydrogen adsorption descriptors, and two simple descriptors, (χ/M1)×(r/M2)2 and (r/M2)3(m/X1), were identified to be applied to chemisorption and physical adsorption, respectively. The results could provide a theoretical basis for the subsequent synthesis of MXene materials with excellent hydrogen storage properties.

Multi-objective optimization design of recycled aggregate concrete mixture proportions based on machine learning and NSGA-II algorithm

This paper employs Support Vector Regression (SVR), Random Forest Regression (RF), Gradient Boosting (GB), and Extreme Gradient Boosting (XGB) algorithms to establish the compressive strength prediction models for Recycled Aggregate Concrete (RAC) and analyze the influence of ten inputs on RAC compressive strength. Combined with the best prediction model, the Non-dominated Sorting Genetic Algorithm Ⅱ (NSGA-II) is applied for multi-objective optimization of mixture proportions in RAC addressing cost, carbon emissions, and compressive strength as key objectives. The results demonstrate that the RAC compressive strength prediction model using the GB algorithm exhibits the highest accuracy, with the R2 value of 0.99 for the training set and 0.85 for the testing set. Feature importance and SHAP analysis reveal that cement and water contents are the primary factors affecting the compressive strength of RAC. Among the three mineral admixtures including Silica Fume (SF), Fly Ash (FA), and Ground Granulated Blast Furnace Slag (GGBFS), SF exhibits superior improvement in the compressive strength of RAC compared to FA and GGBFS. Partial dependency plot analysis indicates that concrete strength remains unaffected by recycled coarse aggregate content within the range of 0 to 300 kg/m3. Pareto fronts of a tri-objective mixture optimization problem for RAC are successfully obtained through the GB model combined with the NSGA-II algorithm, which can guide the optimization of RAC preparation.

Machine learning approach for satellite-based subfield canola yield prediction using floral phenology metrics and soil parameters

Early monitoring of within-field yield variability and forecasting yield potential is critical for farmers and other key stakeholders such as policymakers. Remote sensing techniques are progressively being used in yield prediction studies due to easy access and affordability. Despite the increasing use of remote sensing techniques for yield prediction in agriculture, there is still a need for medium-resolution satellite imagery when predicting canola yield using a combination of crop and soil information. In this study, we investigated the utility of remotely sensed flowering information from PlanetScope (at 4 m) satellite imagery combined with derived soil and topography parameters to predict canola yield. Our yield prediction model was trained and validated using data from 21 fields managed under variable rate seed and fertilizer application, including cleaned harvester yield maps, soil, and topography maps. To quantify the flowering intensity of canola, 9 vegetation indices (VIs) were calculated using spectral bands from PlanetScope imagery acquired for the reproductive stages of canola. We created five random forest regression models using different subsets of covariates, including VIs, soil, and topography features, to predict canola yield within the season. Using a random forest regression algorithm, we recorded accuracies ranging from poor to best performing using coefficient of determination and root mean squared error (R2: 0.47 to 0.66, RMSE: 325 to 399 kg ha−1). The optimal subset of covariates identified electrical conductivity (EC), Normalized Difference Yellowness Index, and Canola Index as the key variables explaining within-spatial variability in canola yield. Our final model exhibited a validation R2 of 0.46 (RMSE = 730 kg ha−1), demonstrating the potential of medium-resolution satellite imagery during the flowering stage to detect and quantify sub-field spatial and temporal floral phenology changes when predicting canola yield.

Study on enhancing waste PVC management through predictive Machine Learning analysis of TGA and its economic benefits

The onset of COVID-19 has led to a sudden surge of existing medical plastic wastes, majority of them made up of PVCs. Pyrolysis, a globally adopted waste management alternative, can be a crucial aid to handle the bulk plastic wastes which makes the identification of thermal behaviour of materials important to aid in the scale-up of pyrolytic processes. The study presents a novel approach of utilising regression-based algorithms- Support Vector Regression (SVR), Random Forest Regression (RFR) and K-Nearest Neighbour algorithms (KNN), to predict the weight loss % at a heating rate of 20 K/min for 6 different PVC based medical plastic wastes. The correlation between the factors affecting Thermogravimetric Analysis (TGA) have been identified using heat maps, the Machine Learning (ML) models were trained on this TGA data and the model efficiencies were identified using performance metrics of Mean Squared Error (MSE), Mean Absolute Error (MAE) and R2. Several kinetic parameters like activation energy and reaction order were estimated using Coats - Redfern method followed by a cost analysis. The results showed that experimental and predicted weight loss were in good agreement with a regression R2 value > 0.97 for all the materials except the composite outer cover. Coats - Redfern method was successful in the estimation of kinetic parameters and the economic analysis indicated that the utilisation of ML in TGA analysis can have significant financial benefits for the concerned industry.

Performance Comparison of different Disease Detection using Stacked Ensemble Learning Model

Malignancy risks and genetic disorders have long been challenging due to procedures that lack precision and predictability, thereby complicating the precise identification of diseases and their root causes. Machine learning classifiers have emerged as more suitable and effective tools. Various machine learning classifiers have been utilized to examine different genetic disorders, and the results from these classifiers have been further compared to determine their superiority. In this study, a variety of classifiers, including the SVM, KNN, decision tree, random forest, and logistic regression algorithms, are examined. These classifiers utilize specific training variables to analyze how input values correspond to the respective class. After successfully implementing each classifier, we proceeded to employ Stacking, an ensemble machine learning technique that aggregates predictions from individual classifiers on the same dataset. Four datasets, including the breast cancer, diabetes, Parkinson’s, and genomic datasets, were successfully implemented using the aforementioned methods, and the results obtained showed how the input values correspond to the class using a few training variables. SVM classifier was shown to be the most effective of the five described classifiers, having the highest accuracy in most of the cases. It provided accuracies of 97.43%, 97.46%, 97.45%, and 97.44% for each of the genome cancer, diabetes, Parkinson’s, and breast cancer datasets. The KNN and Random Forest models also came out to be very effective, with accuracy around 95% and 91%, respectively, for various disease datasets. The Logistic Regression and Decision Tree models also worked well. However, the ensemble method of Stacking proved to be highly efficient above all other base models and generated accuracies above 97.5% for all the aforementioned diseases.

Will the real populists please stand up? A machine learning index of party populism

The existing literature on populism has seen numerous attempts to empirically quantify this somewhat ambiguous concept. Despite notable advances, continuous measures of populism with a clear theoretical background and a considerable coverage are still hard to come by. This paper proposes a novel approach to measuring party populism by combining several different expert-surveys via supervised machine learning techniques. Employing the random forest regression algorithm, we greatly expand the geographical and temporal coverage of two well-known populism indicators, which are based on the discursive and the ideational approach, respectively. The resulting multidimensional measures capture party-level populism on a continuous 0–10 scale, covering 1920 parties in 169 countries from 1970 to 2019. Our measures accurately replicate both definitions of populism, although the indicators may be more suitable for predicting populist outcomes in Western countries, as compared to non-Western ones.

Software Effort Estimation using Machine Learning Algorithms

Effort estimation is a crucial aspect of software development, as it helps project managers plan, control, and schedule the development of software systems. This research study compares various machine learning techniques for estimating effort in software development, focusing on the most widely used and recent methods. The paper begins by highlighting the significance of effort estimation and its associated difficulties. It then presents a comprehensive overview of the different categories of effort estimation techniques, including algorithmic, model-based, and expert-based methods. The study concludes by comparing methods for a given software development project. Random Forest Regression algorithm performs well on the given dataset tested along with various Regression algorithms, including Support Vector, Linear, and Decision Tree Regression. Additionally, the research identifies areas for future investigation in software effort estimation, including the requirement for more accurate and reliable methods and the need to address the inherent complexity and uncertainty in software development projects. This paper provides a comprehensive examination of the current state-of-the-art in software effort estimation, serving as a resource for researchers in the field of software engineering.

Forest Canopy Height Retrieval and Analysis Using Random Forest Model with Multi-Source Remote Sensing Integration

Forest canopy height is an important indicator of the forest ecosystem, and an accurate assessment of forest canopy height on a large scale is of great significance for forest resource quantification and carbon sequestration. The retrieval of canopy height based on remote sensing provides a possibility for studying forest ecosystems. This study proposes a new method for estimating forest canopy height based on remote sensing. In this method, the GEDI satellite and ICESat-2 satellite, which are different types of space-borne lidar products, are used to cooperate with the Landsat 9 image and SRTM terrain data, respectively. Two forest canopy height-retrieval models based on multi-source remote sensing integration are obtained using a random forest regression (RFR) algorithm. The study, conducted at a forest site in the northeastern United States, synthesized various remote sensing data sets to produce a robust canopy height model. First, we extracted relative canopy height products, multispectral features, and topographic data from GEDI, ICESat-2, Landsat 9, and SRTM images, respectively. The importance of each variable was assessed, and the random forest algorithm was used to analyze each variable statistically. Then, the random forest regression algorithm was used to combine these variables and construct the forest canopy height model. Validation with airborne laser scanning (ALS) data shows that the GEDI and ICESat-2 models using a single data source achieve better accuracy than the Landsat 9 model. Notably, the combination of GEDI, Landsat 9, and SRTM data (R = 0.92, MAE = 1.91 m, RMSE = 2.78 m, and rRMSE = 12.64%) and a combination of ICESat-2, Landsat 9, and SRTM data (R = 0.89, MAE = 1.84 m, RMSE = 2.54 m, and rRMSE = 10.75%). Compared with the least accurate Landsat 9 model, R increased by 29.58%, 93.48%, MAE by 44.64%, 46.20%, RMSE by 42.80%, 49.40%, and the rRMSE was increased by 42.86% and 49.32%, respectively. These results fully evaluate and discuss the practical performance and benefits of multi-source data retrieval of forest canopy height by combining space-borne lidar data with Landsat 9 data, which is of great significance for understanding forest structure and dynamics. The study provides a reliable methodology for estimating forest canopy height and valuable insights into forest resource management and its contribution to global climate change.