Algorithm 1.
2-Step Optimization Workflow for Data Fusion Process.
1. Input: 2D input maps () for each of the 5 climate features, 2D input maps () compacting all 11 socioeconomic-governmental factors, size of lead-time () for Sub-Scenarios (a-f), values of model's hyperparameters (), threshold values for precision metric (), validation dataset, testing dataset |
2. Output: 2D output maps () representing spatial classes distribution for all 251 countries with defined lead-time using the proposed approach |
3. for each ← Step I of Optimization Process |
do |
a. initialize training of deep learning prediction model with user-defined values & with selected Sub-Scenario (a-f) for defined in Step I optimization process |
b. derive predictions on validation and testing datasets via trained model with defined & values |
c. model evaluation of predictions on testing dataset via score |
if ≥ |
Proceed |
d. store score with corresponding values & type of Sub-Scenario for defined |
4. select optimal with highest score with corresponding values & type of Sub-Scenario for defined |
5. do |
a. fuse optimal with |
b. initialize training of deep learning prediction model with selected values & type of Sub-Scenario for defined from (4) ← Step II of Optimization Process |
c. derive predictions on validation and testing datasets with trained model |
d. model evaluation of predictions on testing dataset via score |
if ≥ |
Proceed |
store best prediction model for near real-time predictions |
Else |
re-train prediction model with additional hyperparameters () tuning |
6. end |