Efficient and Accurate Zero-Day Electricity Theft Detection from Smart Meter Sensor Data Using Prototype and Ensemble Learning

. 2025 Jul 1;25(13):4111. doi: 10.3390/s25134111

Algorithm 3 The Grid Search algorithm for fine-tuning the models.

Input: Vector of n sets of parameters P, for a specific classification algorithm a. Scoring function s that is to be maximized.
Output: Set of parameters p that produces the highest value for s

1:
param_combinations ← CartesianProduct(P)
2:
maximal_score ← −∞
3:
best_params ← None
4:
for p ∈ param_combinations do
5:
$model \leftarrow a (p)$
6:
$score \leftarrow s (model)$
7:
if score > maximal_score then
8:
maximal_score ← score
9:
maximal_params ← p
10:
end if
11:
end for
12:
return best_params