A Neural Network trainer, which can help us observe the training process.
** 1. generate_data.py: generate the training data and test data. This program first generate a sequence of points by using numpy.random.rand function in a rectangle (-1.5, 1.5). The forumlar is as follows: scaled_value = min + (rand_value*(max-min)).
Second, we check every point whether it is in a unit circle and label it with 1, otherwise with 0.
Third, we write all these points and their labels into a csv file.
** 2. data_loader.py: load data from the csv file and translate it into Tensor.
** 3. model.py: define a fully connected neural network model.
** 4. main.py: define a training progress.