Classification of Large-Scale Mobile Laser Scanning Data in Urban Area with LightGBM

Sevgen, Eray; Abdikan, Saygin

doi:10.3390/rs15153787

Cited by 8 publications

(7 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…KPConv, known for its novel convolution operation tailored for point clouds, stands out for its efficiency in semantic segmentation tasks and it shows high performance over several datasets [6,8,73]. LightGBM, a gradient-boosting framework, is chosen for its accuracy and high efficiency, particularly when dealing with handcrafted features from point clouds [74]. 3DMASC, the most recent one, offers an accessible and explainable approach for point cloud classification, managing diverse datasets effectively [75].…”

Section: Baseline Methodsmentioning

confidence: 99%

Multi-Context Point Cloud Dataset and Machine Learning for Railway Semantic Segmentation

Kharroubi,

Ballouch,

Hajji

et al. 2024

Infrastructures

View full text Add to dashboard Cite

Railway scene understanding is crucial for various applications, including autonomous trains, digital twining, and infrastructure change monitoring. However, the development of the latter is constrained by the lack of annotated datasets and limitations of existing algorithms. To address this challenge, we present Rail3D, the first comprehensive dataset for semantic segmentation in railway environments with a comparative analysis. Rail3D encompasses three distinct railway contexts from Hungary, France, and Belgium, capturing a wide range of railway assets and conditions. With over 288 million annotated points, Rail3D surpasses existing datasets in size and diversity, enabling the training of generalizable machine learning models. We conducted a generic classification with nine universal classes (Ground, Vegetation, Rail, Poles, Wires, Signals, Fence, Installation, and Building) and evaluated the performance of three state-of-the-art models: KPConv (Kernel Point Convolution), LightGBM, and Random Forest. The best performing model, a fine-tuned KPConv, achieved a mean Intersection over Union (mIoU) of 86%. While the LightGBM-based method achieved a mIoU of 71%, outperforming Random Forest. This study will benefit infrastructure experts and railway researchers by providing a comprehensive dataset and benchmarks for 3D semantic segmentation. The data and code are publicly available for France and Hungary, with continuous updates based on user feedback.

show abstract

Section: Baseline Methodsmentioning

confidence: 99%

Multi-Context Point Cloud Dataset and Machine Learning for Railway Semantic Segmentation

Kharroubi,

Ballouch,

Hajji

et al. 2024

Infrastructures

View full text Add to dashboard Cite

show abstract

“…Thus, not only the geometric structure of the point cloud is used, but also the color information of the objects. Sevgen and Abdikan [17] presented a study on the classification of mobile LiDAR point clouds with LightGBM using hand-crafted features. Krishna Moorthy et al [18] tested Random forest (RF), XGBoost and lightGBM algorithms with individual tree and field data from tropical and deciduous forests, using geometric features calculated at multiple spatial scales.…”

Section: Related Workmentioning

confidence: 99%

“…In this study, a bagging method named RF and two boosting methods such as LightGBM and XGBoost were selected as classifiers. RF, XGboost and LightGBM have also been applied for point cloud classification in recent years [18], [43], [17].…”

Section: F Machine Learning Classifiersmentioning

confidence: 99%

Explainable Artificial Intelligence for Machine Learning-Based Photogrammetric Point Cloud Classification

Atik,

Duran,

Seker

2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Point clouds are one of the widely used data sources for spatial modeling. Artificial intelligence approaches have become an important tool for understanding and extracting semantic information of point clouds. In particular, the explainability of machine learning approaches for 3D data has not been sufficiently investigated. Moreover, existing studies are generally limited to object classification issues. This is a pioneer study that addresses the classification of photogrammetric point clouds in terms of explainable artificial intelligence. In this study, the explainability of black-box machine learning models in the context of the classification of photogrammetric point clouds was investigated. Each point in the point cloud is defined using geometric and spectral features. Additionally, the effect of selecting the most important of these features on the classification performance of ML models such as Random Forest, XGBoost and LightGBM was examined. The explainability of ML models was analyzed with Shapley Additive exPlanation (SHAP), an explainable artificial intelligence approach. SHAP analysis was compared with filterbased Information Gain (IG) and ReliefF methods for feature selection. Using the features selected with SHAP analysis, overall accuracy (OA) of 85.50% in the Ankeny dataset, 91.70% in the Building dataset and 83.28% in the Cadastre dataset was achieved with LightGBM. The evaluation with XGBoost shows overall accuracy of 85.22% for Ankeny, 91.21% for Building and 82.47% for Cadastre. The evaluation with RF shows overall accuracy of 83.70% for Ankeny, 89.08% for Building and 79.36% for Cadastre.

show abstract

“…This study employed LightGBM, a gradient-boosting framework, for loan approval classification. LightGBM was chosen for its efficiency, effectiveness in handling large datasets, and ability to capture non-linear relationships between features and target variables (Noviandy et al, 2023d;Sevgen & Abdikan, 2023). Additionally, LightGBM's speed and scalability make it well-suited for the large-scale datasets commonly encountered in finance and banking.…”

Section: Lightgbm Modelmentioning

confidence: 99%

Enhancing Loan Approval Decision-Making: An Interpretable Machine Learning Approach Using Lightgbm for Digital Economy Development

Noviandy,

Idroes,

Hardi

2024

MJoC

View full text Add to dashboard Cite

This study aims to enhance loan approval decision-making in the digital economy using an interpretable machine learning approach. The primary research question investigates how integrating an interpretable machine learning approach can improve the accuracy and transparency of loan approval processes. We employed LightGBM, a gradient-boosting framework for loan approval classification, optimized via Random Search hyperparameter tuning and validated using 10-fold cross-validation. We incorporated the Shapley Additive exPlanations (SHAP) framework to address the challenge of interpretability in machine learning. The LightGBM model outperformed conventional algorithms (Decision Tree, Random Forest, AdaBoost, and Extra Trees) in accuracy (98.13%), precision (97.78%), recall (97.17%), and F1-score (97.48%). The study demonstrates that using an interpretable machine learning approach with LightGBM and SHAP can significantly improve the accuracy and transparency of loan approval decisions. This method offers a promising avenue for financial institutions to enhance their loan approval mechanisms, ensuring more reliable, efficient, and transparent decision-making in the digital economy. The study also underscores the importance of interpretability in deploying machine learning solutions in sectors with significant socio-economic impacts.

show abstract

Classification of Large-Scale Mobile Laser Scanning Data in Urban Area with LightGBM

Cited by 8 publications

References 60 publications

Multi-Context Point Cloud Dataset and Machine Learning for Railway Semantic Segmentation

Multi-Context Point Cloud Dataset and Machine Learning for Railway Semantic Segmentation

Explainable Artificial Intelligence for Machine Learning-Based Photogrammetric Point Cloud Classification

Enhancing Loan Approval Decision-Making: An Interpretable Machine Learning Approach Using Lightgbm for Digital Economy Development

Contact Info

Product

Resources

About