In this repository, you will build and validate a classification model using satellite image data.
The dataset provided consists of two .csv
files:
- Image Data: Contains an eight-band satellite image data.
- Labels: Contains the corresponding labels.
The dataset includes the following columns:
- x and y: Coordinates of the pixels.
- band1-band8: Brightness values for each band.
- label: Indicates the land cover type for the pixels.
-
Model Selection and Validation:
- Choose a classification algorithm to build your model.
- Test multiple algorithms.
- Determine which variables to use for splitting the data into training and validation samples.
- Justify your chosen validation approach.
-
Classification Results:
- Display the classification results using a confusion matrix and provide relevant accuracy/error metrics.
-
Visualization:
- Plot the original satellite image
- Plot predicted satellite image coloured according to specified labels