交通流量預(yù)測(cè)在智能交通(ITS)系統(tǒng)中占有重要地位兵拢，是實(shí)現(xiàn)交通誘導(dǎo)的前提翻斟。準(zhǔn)確實(shí)時(shí)的短時(shí)交通流預(yù)測(cè)有助于更好的分析路網(wǎng)交通狀況，對(duì)路網(wǎng)交通規(guī)劃和交通優(yōu)化控制有非常重要的作用说铃。隨著交通數(shù)據(jù)采集技術(shù)的不斷發(fā)展访惜，及時(shí)獲取路網(wǎng)中實(shí)時(shí)的交通數(shù)據(jù)已成為可能嘹履，大量的交通信息為基于深度學(xué)習(xí)的預(yù)測(cè)模型提供了數(shù)據(jù)保障。
基于神經(jīng)網(wǎng)絡(luò)的交通流預(yù)測(cè)的相關(guān)研究如下列論文所示债热。由于與我的研究方向相關(guān)砾嫉，在本文中我實(shí)現(xiàn)了基于SAEs、RNN的交通流量預(yù)測(cè)模型窒篱。

Paper：
Traffic Flow Prediction With Big Data: A Deep Learning Approach
Using LSTM and GRU neural network methods for traffic flow prediction

Github：https://github.com/xiaochus/TrafficFlowPrediction

環(huán)境

• Python 3.5
• Tensorflow-gpu 1.2.0
• Keras 2.1.3
• scikit-learn 0.18

數(shù)據(jù)處理

實(shí)驗(yàn)數(shù)據(jù)是從Caltrans Performance Measurement System (PeMS)獲取的焕刮。原始的流量數(shù)據(jù)是一個(gè)長(zhǎng)度為n的一維數(shù)據(jù)。我們首先使用訓(xùn)練集的數(shù)據(jù)實(shí)現(xiàn)一個(gè)標(biāo)準(zhǔn)化對(duì)象scaler墙杯，然后使用scaler分別對(duì)訓(xùn)練集與測(cè)試集進(jìn)行標(biāo)準(zhǔn)化配并。
由于時(shí)序預(yù)測(cè)任務(wù)需要使用歷史數(shù)據(jù)對(duì)未來(lái)數(shù)據(jù)進(jìn)行預(yù)測(cè)，我們使用時(shí)滯變量lags對(duì)數(shù)據(jù)進(jìn)行劃分高镐，最后獲得大小為(samples, lags)的數(shù)據(jù)集溉旋。
劃分好的數(shù)據(jù)集在排列順序上依舊帶有時(shí)序特性，雖然keras在訓(xùn)練時(shí)可以選擇對(duì)數(shù)據(jù)進(jìn)行混洗嫉髓，但是其執(zhí)行順序是先對(duì)val數(shù)據(jù)進(jìn)行采樣再進(jìn)行混洗观腊，采樣過(guò)程依舊是按照順序來(lái)的。因此我們事先使用np.random.shuffle對(duì)數(shù)據(jù)進(jìn)行混洗岩喷，打亂數(shù)據(jù)的順序恕沫。

def process_data(train, test, lags):
    """Process data
    Reshape and split train\test data.

    # Arguments
        train: String, name of .csv train file.
        test: String, name of .csv test file.
        lags: integer, time lag.
    # Returns
        X_train: ndarray.
        y_train: ndarray.
        X_test: ndarray.
        y_test: ndarray.
        scaler: StandardScaler.
    """
    attr = 'Lane 1 Flow (Veh/5 Minutes)'
    df1 = pd.read_csv(train, encoding='utf-8').fillna(0)
    df2 = pd.read_csv(test, encoding='utf-8').fillna(0)

    # scaler = StandardScaler().fit(df1[attr].values)
    scaler = MinMaxScaler(feature_range=(0, 1)).fit(df1[attr].values)
    flow1 = scaler.transform(df1[attr].values)
    flow2 = scaler.transform(df2[attr].values)

    train, test = [], []
    for i in range(lags, len(flow1)):
        train.append(flow1[i - lags: i + 1])
    for i in range(lags, len(flow2)):
        test.append(flow2[i - lags: i + 1])

    train = np.array(train)
    test = np.array(test)
    np.random.shuffle(train)

    X_train = train[:, :-1]
    y_train = train[:, -1]
    X_test = test[:, :-1]
    y_test = test[:, -1]

    return X_train, y_train, X_test, y_test, scaler

模型

LSTM

2隱層LSTM網(wǎng)絡(luò)。

LSTM.png

def get_lstm(units):
    """LSTM(Long Short-Term Memory)
    Build LSTM Model.

    # Arguments
        units: List(int), number of input, output and hidden units.
    # Returns
        model: Model, nn model.
    """

    model = Sequential()
    model.add(LSTM(units[1], input_shape=(units[0], 1), return_sequences=True))
    model.add(LSTM(units[2]))
    model.add(Dropout(0.2))
    model.add(Dense(units[3], activation='linear'))

    return model

GRU

2隱層GRU網(wǎng)絡(luò)纱意。

GRU.png

def get_gru(units):
    """GRU(Gated Recurrent Unit)
    Build GRU Model.

    # Arguments
        units: List(int), number of input, output and hidden units.
    # Returns
        model: Model, nn model.
    """

    model = Sequential()
    model.add(GRU(units[1], input_shape=(units[0], 1), return_sequences=True))
    model.add(GRU(units[2]))
    model.add(Dropout(0.2))
    model.add(Dense(units[3], activation='linear'))

    return model

SAEs

SAEs.png

Auto-Encoders的原理是先通過(guò)一個(gè)encode層對(duì)輸入進(jìn)行編碼婶溯，這個(gè)編碼就是特征，然后利用encode乘第2層參數(shù)（也可以是encode層的參數(shù)的轉(zhuǎn)置乘特征并加偏執(zhí)）偷霉，重構(gòu)（解碼）輸入迄委，然后用重構(gòu)的輸入和實(shí)際輸入的損失訓(xùn)練參數(shù)。
這里我們構(gòu)建了三個(gè)單獨(dú)的自動(dòng)編碼器类少，并按照相同的隱層結(jié)構(gòu)構(gòu)建了一個(gè)三層的SAEs叙身。

def _get_sae(inputs, hidden, output):
    """SAE(Auto-Encoders)
    Build SAE Model.

    # Arguments
        inputs: Integer, number of input units.
        hidden: Integer, number of hidden units.
        output: Integer, number of output units.
    # Returns
        model: Model, nn model.
    """

    model = Sequential()
    model.add(Dense(hidden, input_dim=inputs, name='hidden'))
    model.add(Activation('sigmoid'))
    model.add(Dropout(0.2))
    model.add(Dense(output))

    return model


def get_saes(layers):
    """SAEs(Stacked Auto-Encoders)
    Build SAEs Model.

    # Arguments
        layers: List(int), number of input, output and hidden units.
    # Returns
        models: List(Model), List of SAE and SAEs.
    """
    sae1 = _get_sae(layers[0], layers[1], layers[-1])
    sae2 = _get_sae(layers[1], layers[2], layers[-1])
    sae3 = _get_sae(layers[2], layers[3], layers[-1])

    saes = Sequential()
    saes.add(Dense(layers[1], input_dim=layers[0], name='hidden1'))
    saes.add(Activation('sigmoid'))
    saes.add(Dense(layers[2], name='hidden2'))
    saes.add(Activation('sigmoid'))
    saes.add(Dense(layers[3], name='hidden3'))
    saes.add(Activation('sigmoid'))
    saes.add(Dropout(0.2))
    saes.add(Dense(layers[4]))

    models = [sae1, sae2, sae3, saes]

    return models

訓(xùn)練

LSTM、GRU按照正常的RNN網(wǎng)絡(luò)進(jìn)行訓(xùn)練硫狞。使用train_model()函數(shù)訓(xùn)練信轿。
SAEs的訓(xùn)練過(guò)程：多個(gè)SAE分別訓(xùn)練，第一個(gè)SAE訓(xùn)練完之后残吩，其encode的輸出作為第二個(gè)SAE的輸入财忽，接著訓(xùn)練。最后訓(xùn)練完后泣侮，將所有SAE的中間隱層連接起來(lái)組成一個(gè)SAEs網(wǎng)絡(luò)即彪，使用之前的權(quán)值作為初始化權(quán)值，再對(duì)整個(gè)網(wǎng)絡(luò)進(jìn)行fine-tune活尊。使用train_seas()函數(shù)訓(xùn)練隶校。

使用RMSprop(lr=0.001, rho=0.9, epsilon=1e-06)作為優(yōu)化器漏益，batch_szie為256，lags為12(即時(shí)滯長(zhǎng)度為一個(gè)小時(shí))深胳。

def train_model(model, X_train, y_train, name, config):
    """train
    train a single model.

    # Arguments
        model: Model, NN model to train.
        X_train: ndarray(number, lags), Input data for train.
        y_train: ndarray(number, ), result data for train.
        name: String, name of model.
        config: Dict, parameter for train.
    """

    model.compile(loss="mse", optimizer="rmsprop", metrics=['mape'])
    # early = EarlyStopping(monitor='val_loss', patience=30, verbose=0, mode='auto')
    hist = model.fit(
        X_train, y_train,
        batch_size=config["batch"],
        epochs=config["epochs"],
        validation_split=0.05)

    model.save('model/' + name + '.h5')
    df = pd.DataFrame.from_dict(hist.history)
    df.to_csv('model/' + name + ' loss.csv', encoding='utf-8', index=False)


def train_seas(models, X_train, y_train, name, config):
    """train
    train the SAEs model.

    # Arguments
        models: List, list of SAE model.
        X_train: ndarray(number, lags), Input data for train.
        y_train: ndarray(number, ), result data for train.
        name: String, name of model.
        config: Dict, parameter for train.
    """

    temp = X_train
    # early = EarlyStopping(monitor='val_loss', patience=30, verbose=0, mode='auto')

    for i in range(len(models) - 1):
        if i > 0:
            p = models[i - 1]
            hidden_layer_model = Model(input=p.input,
                                       output=p.get_layer('hidden').output)
            temp = hidden_layer_model.predict(temp)

        m = models[i]
        m.compile(loss="mse", optimizer="rmsprop", metrics=['mape'])
        m.fit(temp, y_train, batch_size=config["batch"],
              epochs=config["epochs"],
              validation_split=0.05)

        models[i] = m

    saes = models[-1]
    for i in range(len(models) - 1):
        weights = models[i].get_layer('hidden').get_weights()
        saes.get_layer('hidden%d' % (i + 1)).set_weights(weights)

    train_model(saes, X_train, y_train, name, config)

實(shí)驗(yàn)

評(píng)估

在這里使用MAE绰疤、MSE、RMSE稠屠、MAPE峦睡、R2、explained_variance_score幾個(gè)指標(biāo)對(duì)回歸預(yù)測(cè)結(jié)果進(jìn)行評(píng)估权埠。

def MAPE(y_true, y_pred):
    """Mean Absolute Percentage Error
    Calculate the mape.

    # Arguments
        y_true: List/ndarray, ture data.
        y_pred: List/ndarray, predicted data.
    # Returns
        mape: Double, result data for train.
    """

    y = [x for x in y_true if x > 0]
    y_pred = [y_pred[i] for i in range(len(y_true)) if y_true[i] > 0]

    num = len(y_pred)
    sums = 0

    for i in range(num):
        tmp = abs(y[i] - y_pred[i]) / y[i]
        sums += tmp

    mape = sums * (100 / num)

    return mape


def eva_regress(y_true, y_pred):
    """Evaluation
    evaluate the predicted resul.

    # Arguments
        y_true: List/ndarray, ture data.
        y_pred: List/ndarray, predicted data.
    """

    mape = MAPE(y_true, y_pred)
    vs = metrics.explained_variance_score(y_true, y_pred)
    mae = metrics.mean_absolute_error(y_true, y_pred)
    mse = metrics.mean_squared_error(y_true, y_pred)
    r2 = metrics.r2_score(y_true, y_pred)
    print('explained_variance_score:%f' % vs)
    print('mape:%f%%' % mape)
    print('mae:%f' % mae)
    print('mse:%f' % mse)
    print('rmse:%f' % math.sqrt(mse))
    print('r2:%f' % r2)

預(yù)測(cè)

我們使用訓(xùn)練好的模型對(duì)測(cè)試集進(jìn)行預(yù)測(cè)榨了。

def main():
    lstm = load_model('model/lstm.h5')
    gru = load_model('model/gru.h5')
    saes = load_model('model/saes.h5')
    models = [lstm, gru, saes]
    names = ['LSTM', 'GRU', 'SAEs']

    lag = 12
    file1 = 'data/train.csv'
    file2 = 'data/test.csv'
    _, _, X_test, y_test, scaler = process_data(file1, file2, lag)
    y_test = scaler.inverse_transform(y_test)

    y_preds = []
    for name, model in zip(names, models):
        if name == 'SAEs':
            X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1]))
        else:
            X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1], 1))
        file = 'images/' + name + '.png'
        plot_model(model, to_file=file, show_shapes=True)
        predicted = model.predict(X_test)
        predicted = scaler.inverse_transform(predicted)
        y_preds.append(predicted[:300])
        print(name)
        eva_regress(y_test, predicted)

    plot_results(y_test[: 300], y_preds, names)

預(yù)測(cè)精度對(duì)比如下所示：

預(yù)測(cè)結(jié)果對(duì)比如下所示：

eva.png