using Grid search algorithm with cross validation to tune parameters and training for various classification models and compare between them this research aplied on digits by sklearn datasets
1-svm (C,kernel) 2-random forest (n_estimators) 3-logistic regression (C) 4-naive_bayes.GaussianNB (var_smoothing) 5-naive_bayes.MultinomialNB (alpha) 6-DecisionTree (criterion)
and used 5 folds for cross validation
1 svm scored:94.7697% with parameters values ={'C': 1, 'kernel': 'linear'}
2 random_forest scored:93.9362% with parameters values ={'n_estimators': 100}
3 logistic_regression scored:92.2114% with parameters values ={'C': 1}
4 nbm scored:87.0350% with parameters values ={'alpha': 0}
5 dt scored:81.0258% with parameters values ={'criterion': 'entropy'}
6 nbg scored:80.6928% with parameters values ={'var_smoothing': 1e-09}