To prepare a Receiver Operating Characteristic (ROC) curve in Origin software for validating a Landslide Susceptibility Model, you can follow these steps:
Prepare Data:Calculate True Positive Rate (TPR) and False Positive Rate (FPR) at different thresholds. Organize data into two columns: FPR in column A, TPR in column B.
Enter Data in Origin:Open a new workbook and input FPR and TPR values.
Create ROC Curve:Highlight both columns. Go to Plot > Basic 2D > Line to create the ROC curve.
Calculate AUC:Use Analysis > Mathematics > Integration... to calculate the Area Under the Curve (AUC).
Customize and Export:Customize axes and labels. Export the graph as an image or PDF.
This will give you a visual and numerical validation of your Landslide Susceptibility Model.
The dataset should include data points that represent conditions leading to landslides (or not) based on various features. Here’s how you can prepare it:
a. Collect Data
b. Data Labeling
Label the data points as:
Positive (1): Locations where landslides occurred.
Negative (0): Locations where landslides did not occur.
c. Data Splitting
Split the dataset into training and testing sets, typically in a 70-30 or 80-20 ratio.
2. Training the Model
Use a machine learning model to predict landslide occurrences based on your features (e.g., logistic regression, random forest, neural networks).
a. Feature Selection
Choose relevant features that contribute to the occurrence of landslides.
b. Model Training
Train your model on the training set using the selected features.
3. Making Predictions
Use the trained model to predict the likelihood of landslides on the test dataset.
4. Computing TPR & FPR
Once you have the predictions, you can compute the True Positive Rate (TPR) and False Positive Rate (FPR).
TPR=TPTP+FN\text{TPR} = \frac{TP}{TP + FN}TPR=TP+FNTPTPR (True Positive Rate): Also known as sensitivity or recall, it is the proportion of actual positives that are correctly identified.
FPR=FPFP+TN\text{FPR} = \frac{FP}{FP + TN}FPR=FP+TNFPFPR (False Positive Rate): It is the proportion of actual negatives that are incorrectly identified as positives.
c. Calculate TPR & FPR
Using your predictions and the actual labels from the test set, you can calculate TP, FN, FP, and TN, and then use the formulas above to compute TPR and FPR.
5. Example Computation
Suppose you have the following confusion matrix after making predictions: