LSTM validation accuracy fluctuating

Question

I am using LSTM to model time series data. My target variable is categorical so I am using one-hot encoding. The goal is to predict the target class based on the given time. My dataset spans over eight days.

   input_nodes = look_back = 10
   batch_size = 128
   train_generator = create_data_generator(train, look_back, outputs, batch_size, class_weights_dict)
   validation_generator = create_test_generator(test, look_back, outputs, batch_size)
model = Sequential()
   model.add(LSTM(50, input_shape=(input_nodes, outputs)))
   #model.add(Dense(50, activation='relu'))
   model.add(Dropout(0.5))
   model.add(Dense(outputs, activation='softmax'))
   model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
fit network
history = model.fit_generator(train_generator,
                                 steps_per_epoch=math.ceil(len(train)/batch_size),
                                 epochs=25,
                                 validation_data=validation_generator,
                                 validation_steps=math.ceil(len(test)/batch_size))

Here are my results.

Is there anything that I can try to improve my validation accuracy and loss?