Perform k-fold cross-validation for tuning the following parameters of a random forest model.
- ntree: 10
- mtry: 75
- maxnodes: 20
k_fold(k, './data/Archaeal_tfpssm.csv', 'performance.csv')- Divide the data into k parts, the number of parts used by each data set
- (training, validation, testing) = (k-2, 1, 1)
- The following shows the example of the 5-fold cross validation.
📁 Archaeal_tfpssm.csv download
This CSV doesn't contain a header. The information of columns as below:
-
V2: labels of proteins- CP: Cytoplasmic
- CW: Cell Wall
- EC: Extracellular
- IM: Inner membrane
-
V3 ~ V5602: the gapped-dipeptide features of each protein
- accuracy = P/N, average of k-fold cross-validation
| set | training | validation | test |
|---|---|---|---|
| fold1 | 0.93 | 0.91 | 0.88 |
| fold2 | 0.92 | 0.91 | 0.89 |
| fold3 | 0.94 | 0.92 | 0.90 |
| fold4 | 0.91 | 0.89 | 0.87 |
| fold5 | 0.90 | 0.92 | 0.87 |
| ave. | 0.92 | 0.91 | 0.88 |
library(randomForest)
k_fold <- function(fold, input_file, output_file){
# model using random forest & tune best parameters
model <- randomForest(ntree, mtry, maxnodes)
# make confusion matrix tabel
resultframe <- data.frame(truth=tmp$V2,
pred=predict(model, type="class"))
# output the confusion matrix
write.csv()
return (your_model)
}Please list the code and its reference.
If needed, you should explain the details, i.e., comment like # ChatGPT, respond to “your prompt,” February 16, 2023.
Data Set:
- Chang, J.-M. M. et al. (2013) Efficient and interpretable prediction of protein functional classes by correspondence analysis and compact set relations. PLoS ONE 8, e75542.
- Chang J-M, Su EC-Y, Lo A, Chiu H-S, Sung T-Y, & Hsu W-L (2008) PSLDoc: Protein subcellular localization prediction based on gapped-dipeptides and probabilistic latent semantic analysis. Proteins: Structure, Function, and Bioinformatics 72(2):693-710.
Code: