Uncertainty and Overfitting in Fluvial Landform Classification Using Laser Scanned Data and Machine Learning: A Comparison of Pixel and Object-Based Approaches.
Zsuzsanna SzabóTomás MikitaGábor NégyesiOrsolya Gyöngyi VargaPéter BuraiLászló Takács-SzilágyiSzilárd SzabóPublished in: Remote. Sens. (2020)
Keyphrases
- machine learning
- data sets
- data analysis
- decision trees
- training data
- data structure
- pattern recognition
- satellite images
- missing data
- image data
- data collection
- preprocessing steps
- database
- machine learning approaches
- high dimensional data
- statistical analysis
- feature space
- uncertain data
- machine learning algorithms
- data quality
- cross validation
- classification method
- machine learning methods
- support vector
- text classification
- active learning
- learning tasks
- classification rules
- data sources
- classification accuracy
- knowledge discovery
- model selection
- edge detection