casiafan / time_seires_prediction_using_lstm

Time series prediction using LSTM classifier

trainingset_selection.py is used to format input data to training data set. It will return files of those enterprises satisfying the screening rule.

neural_network_run.py is used to train neural network for time series prediction. It returns a lstm model and evaluation results of the model.

preprocessing.py provided some functions required in neural_network_run.py, training_set_selection.py and dataset_merging.py.

optional script

dataset_merging.py is used to merge similar enterprises data set (or enterprises whose data is within similar range) based on Q-Q plot and ks-statistic. It will return merged cluster data set files and summary file to illustrate enterprises in each cluster.

time_seires_prediction_using_lstm's Issues

等我中了1000W来关

RT

Can you show me the train data format?

I use my data to train a model which is always have some problems: raise ValueError("No files selected in current id range. Please check the input training set directory, "
ValueError: No files selected in current id range. Please check the input training set directory, input enterprise id range or file format which should be '[0-9]+.csv'