3. Classify the particles using the random forest model - Romain-Laine/MiLeSIM GitHub Wiki

Performed by: ML_PredictWithModel_RF_BATCH.m

Input: This assumes that the output of ExtractObjects_fromImage_BATCH.m and that of ExtractDescriptors_BATCH.m are present in the folders along with the reconstructed data. The trained model needs to be selected.

Paramaters: The size of the border for the saved images can be set. A minimum probability of classification can be set to divert to unknown objects that were classified with low probability.

Output: The code generates an excel spreadsheet for each dataset in the folder where each identified particles is associated with its predicted class. The classified particles are also saved independently with their masks for display and analysis purposes.