import json import logging import os import numpy as np import pandas as pd import tensorflow as tf from keras.callbacks import ModelCheckpoint, CSVLogger, EarlyStopping from keras.models import load_model from keras.utils import np_utils from sklearn.utils import class_weight import arguments import dataset import hyperband import models # create logger import visualize from dataset import load_or_generate_h5data from utils import exists_or_make_path logger = logging.getLogger('logger') logger.setLevel(logging.DEBUG) # create console handler and set level to debug ch = logging.StreamHandler() ch.setLevel(logging.DEBUG) # create formatter formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s') # add formatter to ch ch.setFormatter(formatter) # add ch to logger logger.addHandler(ch) ch = logging.FileHandler("info.log") ch.setLevel(logging.DEBUG) # create formatter formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s') # add formatter to ch ch.setFormatter(formatter) # add ch to logger logger.addHandler(ch) args = arguments.parse() if args.gpu: config = tf.ConfigProto(log_device_placement=True) config.gpu_options.per_process_gpu_memory_fraction = 0.5 config.gpu_options.allow_growth = True session = tf.Session(config=config) def main_paul_best(): char_dict = dataset.get_character_dict() pauls_best_params = models.pauls_networks.best_config pauls_best_params["vocab_size"] = len(char_dict) + 1 main_train(pauls_best_params) def main_hyperband(): char_dict = dataset.get_character_dict() params = { # static params "type": ["paul"], "batch_size": [args.batch_size], "vocab_size": [len(char_dict) + 1], "window_size": [10], "domain_length": [40], "flow_features": [3], "input_length": [40], # model params "embedding_size": [8, 16, 32, 64, 128, 256], "filter_embedding": [8, 16, 32, 64, 128, 256], "kernel_embedding": [1, 3, 5, 7, 9], "hidden_embedding": [8, 16, 32, 64, 128, 256], "dropout": [0.5], "domain_features": [8, 16, 32, 64, 128, 256], "filter_main": [8, 16, 32, 64, 128, 256], "kernels_main": [1, 3, 5, 7, 9], "dense_main": [8, 16, 32, 64, 128, 256], } logger.info("create training dataset") domain_tr, flow_tr, client_tr, server_tr = load_or_generate_h5data(args.train_h5data, args.train_data, args.domain_length, args.window) hp = hyperband.Hyperband(params, [domain_tr, flow_tr], [client_tr, server_tr]) results = hp.run() json.dump(results, open("hyperband.json")) def get_custom_class_weights(client_tr, server_tr): client = client_tr.value.argmax(1) if type(client_tr) != np.ndarray else client_tr.argmax(1) server = server_tr.value.argmax(1) if type(server_tr) != np.ndarray else server_tr.argmax(1) client_class_weight = class_weight.compute_class_weight('balanced', np.unique(client), client) server_class_weight = class_weight.compute_class_weight('balanced', np.unique(server), server) return { "client": client_class_weight, "server": server_class_weight } def main_train(param=None): exists_or_make_path(args.model_path) char_dict = dataset.get_character_dict() domain_tr, flow_tr, client_tr, server_windows_tr = load_or_generate_h5data(args.train_h5data, args.train_data, args.domain_length, args.window) # parameter p = { "type": args.model_type, "batch_size": 64, "window_size": args.window, "domain_length": args.domain_length, "flow_features": 3, "vocab_size": len(char_dict) + 1, # 'dropout': 0.5, 'domain_features': args.domain_embedding, 'embedding_size': args.embedding, 'filter_main': 64, 'flow_features': 3, # 'dense_main': 512, 'dense_main': 64, 'filter_embedding': args.hidden_char_dims, 'hidden_embedding': args.domain_embedding, 'kernel_embedding': 3, 'kernels_main': 3, 'input_length': 40 } if not param: param = p embedding, model, _ = models.get_models_by_params(param) embedding.summary() model.summary() logger.info("define callbacks") callbacks = [] callbacks.append(ModelCheckpoint(filepath=args.clf_model, monitor='val_loss', verbose=False, save_best_only=True)) callbacks.append(CSVLogger(args.train_log)) if args.stop_early: callbacks.append(EarlyStopping(monitor='val_loss', patience=5, verbose=False)) logger.info("compile model") custom_metrics = models.get_metric_functions() model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'] + custom_metrics) server_tr = np_utils.to_categorical(np.max(server_windows_tr, axis=1), 2) if args.class_weights: logger.info("class weights: compute custom weights") custom_class_weights = get_custom_class_weights(client_tr, server_tr) logger.info(custom_class_weights) else: logger.info("class weights: set default") custom_class_weights = None logger.info("start training") model.fit([domain_tr, flow_tr], [client_tr, server_tr], batch_size=args.batch_size, epochs=args.epochs, callbacks=callbacks, shuffle=True, validation_split=0.2, class_weight=custom_class_weights) logger.info("save embedding") embedding.save(args.embedding_model) def main_test(): domain_val, flow_val, client_val, server_val = load_or_generate_h5data(args.test_h5data, args.test_data, args.domain_length, args.window) clf = load_model(args.clf_model, custom_objects=models.get_metrics()) # stats = clf.evaluate([domain_val, flow_val], # [client_val, server_val], # batch_size=args.batch_size) y_pred = clf.predict([domain_val, flow_val], batch_size=args.batch_size, verbose=1) np.save(args.future_prediction, y_pred) char_dict = dataset.get_character_dict() user_flow_df = dataset.get_user_flow_data(args.test_data) domains = user_flow_df.domain.unique()[:-1] def get_domain_features_reduced(d): return dataset.get_domain_features(d[0], char_dict, args.domain_length) domain_features = [] for ds in domains: domain_features.append(np.apply_along_axis(get_domain_features_reduced, 2, np.atleast_3d(ds))) model = load_model(args.embedding_model) domain_features = np.stack(domain_features).reshape((-1, 40)) pred = model.predict(domain_features, batch_size=args.batch_size, verbose=1) np.save("/tmp/rk/domains.npy", domains) np.save("/tmp/rk/domain_features.npy", domain_features) np.save("/tmp/rk/domain_embd.npy", pred) def main_new_model(): exists_or_make_path(args.model_path) char_dict = dataset.get_character_dict() domain_tr, flow_tr, client_tr, server_windows_tr = load_or_generate_h5data(args.train_h5data, args.train_data, args.domain_length, args.window) # parameter p = { "type": args.model_type, "batch_size": 64, "window_size": args.window, "domain_length": args.domain_length, "flow_features": 3, "vocab_size": len(char_dict) + 1, # 'dropout': 0.5, 'domain_features': args.domain_embedding, 'embedding_size': args.embedding, 'filter_main': 64, 'flow_features': 3, # 'dense_main': 512, 'dense_main': 64, 'filter_embedding': args.hidden_char_dims, 'hidden_embedding': args.domain_embedding, 'kernel_embedding': 3, 'kernels_main': 3, 'input_length': 40 } embedding, _, model = models.get_models_by_params(p) embedding.summary() model.summary() logger.info("define callbacks") callbacks = [] callbacks.append(ModelCheckpoint(filepath=args.clf_model, monitor='val_loss', verbose=False, save_best_only=True)) callbacks.append(CSVLogger(args.train_log)) if args.stop_early: callbacks.append(EarlyStopping(monitor='val_loss', patience=5, verbose=False)) logger.info("compile model") custom_metrics = models.get_metric_functions() model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'] + custom_metrics) server_tr = np_utils.to_categorical(np.max(server_windows_tr, axis=1), 2) if args.class_weights: logger.info("class weights: compute custom weights") custom_class_weights = get_custom_class_weights(client_tr, server_tr) logger.info(custom_class_weights) else: logger.info("class weights: set default") custom_class_weights = None logger.info("start training") server_tr = np.stack(np_utils.to_categorical(s, 2) for s in server_windows_tr) model.fit([domain_tr, flow_tr], [client_tr, server_tr], batch_size=args.batch_size, epochs=args.epochs, callbacks=callbacks, shuffle=True, validation_split=0.2, class_weight=custom_class_weights) logger.info("save embedding") embedding.save(args.embedding_model) def main_embedding(): model = load_model(args.embedding_model) domain_encs, labels = dataset.load_or_generate_domains(args.train_data, args.domain_length) domain_embedding = model.predict(domain_encs, batch_size=args.batch_size, verbose=1) visualize.plot_embedding(domain_embedding, labels, path="results/pp3/embd.png") def main_visualization(): domain_val, flow_val, client_val, server_val = load_or_generate_h5data(args.test_h5data, args.test_data, args.domain_length, args.window) logger.info("plot model") model = load_model(args.clf_model, custom_objects=models.get_metrics()) visualize.plot_model(model, os.path.join(args.model_path, "model.png")) try: logger.info("plot training curve") logs = pd.read_csv(args.train_log) visualize.plot_training_curve(logs, "client", "{}/client_train.png".format(args.model_path)) visualize.plot_training_curve(logs, "server", "{}/server_train.png".format(args.model_path)) except Exception as e: logger.warning(f"could not generate training curves: {e}") client_pred, server_pred = np.load(args.future_prediction) logger.info("plot pr curve") visualize.plot_precision_recall(client_val.value, client_pred, "{}/client_prc.png".format(args.model_path)) visualize.plot_precision_recall(server_val.value, server_pred, "{}/server_prc.png".format(args.model_path)) visualize.plot_precision_recall_curves(client_val.value, client_pred, "{}/client_prc2.png".format(args.model_path)) visualize.plot_precision_recall_curves(server_val.value, server_pred, "{}/server_prc2.png".format(args.model_path)) logger.info("plot roc curve") visualize.plot_roc_curve(client_val.value, client_pred, "{}/client_roc.png".format(args.model_path)) visualize.plot_roc_curve(server_val.value, server_pred, "{}/server_roc.png".format(args.model_path)) visualize.plot_confusion_matrix(client_val.value.argmax(1), client_pred.argmax(1), "{}/client_cov.png".format(args.model_path), normalize=False, title="Client Confusion Matrix") visualize.plot_confusion_matrix(server_val.value.argmax(1), server_pred.argmax(1), "{}/server_cov.png".format(args.model_path), normalize=False, title="Server Confusion Matrix") def main_score(): # mask = dataset.load_mask_eval(args.data, args.test_image) # pred = np.load(args.pred) # visualize.score_model(mask, pred) pass def main_data(): char_dict = dataset.get_character_dict() user_flow_df = dataset.get_user_flow_data(args.train_data) logger.info("create training dataset") domain_tr, flow_tr, client_tr, server_tr, _ = dataset.create_dataset_from_flows(user_flow_df, char_dict, max_len=args.domain_length, window_size=args.window) print(f"domain shape {domain_tr.shape}") print(f"flow shape {flow_tr.shape}") print(f"client shape {client_tr.shape}") print(f"server shape {server_tr.shape}") def main(): if "train" in args.modes: main_train() if "hyperband" in args.modes: main_hyperband() if "test" in args.modes: main_test() if "fancy" in args.modes: main_visualization() if "score" in args.modes: main_score() if "paul" in args.modes: main_paul_best() if "data" in args.modes: main_data() if "train_new" in args.modes: main_new_model() if __name__ == "__main__": main()