Android TensorFlow (with wifi & refractoring changes)

[3ygun]

Goal

Update the Android app to allow for a TensorFlow back-end #23 with the wifi & refracting changes of #27. The validity of which was checked in #11.

Status

NOTE: Don't merge until all are complete!

• [x] Add TensorFlow trainer in addition to custom CrowdML Trainers
• [x] Merge in #27 with Wifi & refractoring changes
• [x] Validate working of:
  • [x] TensorFlow trainer
    • Currently, seems to not get weight updates between batches
  • [x] Regular trainer
    • Is it even currently accessible?
• [x] Remove old commented code in BackgroundDataSend.java

End of Scope for this PR

• [x] Working TensorFlow implementation on the current 1 layer NN (mnist_mlp.pb file) which uses the mnsit 50 data
• [x] Ability to switch back to the old implementations
• NOT: Fully configurable TensorFlow graphs yet #29

Other Parts

• [x] Updated to Gradle 3.3 which seemed to mitigate build errors on my end
• [x] Removed/.gitignored more configuration files

[3ygun]

Should be good I think @tylermzeller could you look over it?

[3ygun]

I tested with the included server/config/tensorflow.json and the mnist_mlp.pd file generated by the following:

import tensorflow as tf

with tf.Session() as sess:

    x = tf.placeholder(tf.float32, shape=[None, 50], name="x")
    y = tf.placeholder(tf.float32, [None, 10], name="y")
    w = tf.placeholder(tf.float32, [50, 10], name="weights_in")

    W = tf.Variable(tf.zeros([50, 10]), name="weights")
    b = tf.Variable(tf.zeros([10]))

    y_out = tf.add(tf.matmul(x, W), b, name="y_out")

    #cross_entropy = tf.reduce_mean(-tf.reduce_sum(y * tf.log(y_out), reduction_indices=[1]))
    cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=y_out))
    train_step = tf.train.AdamOptimizer(0.005).minimize(cross_entropy, name="train")

    correct_prediction = tf.equal(tf.argmax(y_out,1), tf.argmax(y,1))
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32),                                                                                                                                           name="test")

    init = tf.variables_initializer(tf.global_variables(), name="init")

    tf.train.write_graph(sess.graph_def,
                         './',
                         'mnist_mlp.pb', as_text=False)