Easy to use feed forward neural network
Instructions:
to run, navigate to src and run the python script command_line_tools.py
to perform a specific task, enter: task_name --option_1=.. and so on so forth for all options
for example, to train:
train --train_data=../my_train_data.csv epochs=50 test_data../my_test_data.csv hlayers=14,12 classes=6 learn_rate0.01 outfile=../results.csv
Note that ALL options for the required task must be filled in
Training options:
train_data: a csv file containing training data. The first column of each line should be the classification, with the rest being features
epochs: the number of epochs to train for
test_data: a csv file containing data to test for. Test results are generated after each epoch to compare to training results at the same time. Same format as train data
hlayers: hidden layer configuration of network. List of numbers, no spaces. Each entry in the list contains the number of neurons the network contains in the respective hidden layer, going in order from closes to input to closes to output.
classes: number of possible classifications for the network
learn_rate: the learning rate to use during gradient descent
Predict options:
infile: input file to predict with. Format should be the same as other csv files such as train_data and test_data
outifle: file in which the results of the predictions are stored
Save options:
save_dir: directory in which to save the weights of the neural network as they currently are
Load options:
load_dir directory from which to load saved neural network files\