# import tf for classification import tensorflow as tf from load_data import LoadData # make a data loader obj loadData = LoadData("./kr-vs-k.csv") # retrive the data dict data = loadData.load_processed_data() # a input function def input_fn_generator(data): def input_fn(): # load the features features = data["features"].values labels = data["labels"].values # convert the data into tensors features = tf.convert_to_tensor(value=features, dtype=tf.float32) labels = tf.convert_to_tensor(value=labels, dtype=tf.int32) # make a tensor dataset dataset = tf.data.Dataset.from_tensor_slices((features, labels)) # make a batch of data dataset = dataset.shuffle(100).batch(64) # make a oneshot iterator iterator = tf.compat.v1.data.make_one_shot_iterator(dataset)
print("Saving model {}".format(learner.__class__.__name__)) save_model(learner, learner.__class__.__name__) # predict model print("Predicting Model") prediction_test = learner.predict(X_test) results["accu"] = f1_score(y_test, prediction_test, average="micro") return results # load data dataLoader = LoadData("./kr-vs-k.csv") # data loader for kr-vs-k data = dataLoader.load_processed_data() # load preprocessed data # split data X_train, X_test, y_train, y_test = train_test_split(data["features"], data["labels"], test_size=0.2, random_state=42) # initalize classifiers clf_DT = DecisionTreeClassifier() # decission tree classifier clf_KNN = KNeighborsClassifier(n_neighbors=4) # KNN classifier clf_RF = RandomForestClassifier( n_estimators=1000) # Random Forest with 100 trees # list of classifiers to iterate classifier_list = [clf_DT, clf_KNN, clf_RF]