deep learning学习笔记---MemN2N

1. Summary

MemN2N is a generalization of RNN
1) The sentence in MemN2N is equivalent to the word in RNN;

2. Kernel Code

Build Model

def build_model(self):

 self.build_memory()

 self.W = tf.Variable(tf.random_normal([self.edim, self.nwords], stddev=self.init_std))

 z = tf.matmul(self.hid[-1], self.W)

 self.loss = tf.nn.softmax_cross_entropy_with_logits(logits=z, labels=self.target)

 self.lr = tf.Variable(self.current_lr)

 self.opt = tf.train.GradientDescentOptimizer(self.lr)

 params = [self.A, self.B, self.C, self.T_A, self.T_B, self.W]

 grads_and_vars = self.opt.compute_gradients(self.loss,params)

 clipped_grads_and_vars = [(tf.clip_by_norm(gv[0], self.max_grad_norm), gv[1]) for gv in grads_and_vars]

 inc = self.global_step.assign_add(1)

 with tf.control_dependencies([inc]):

 self.optim = self.opt.apply_gradients(clipped_grads_and_vars)

 tf.global_variables_initializer().run()

 self.saver = tf.train.Saver()

Build Memory

def build_memory(self):

 self.global_step = tf.Variable(0, name="global_step")

 self.A = tf.Variable(tf.random_normal([self.nwords, self.edim], stddev=self.init_std))

 self.B = tf.Variable(tf.random_normal([self.nwords, self.edim], stddev=self.init_std))

 self.C = tf.Variable(tf.random_normal([self.edim, self.edim], stddev=self.init_std))

 # Temporal Encoding

 self.T_A = tf.Variable(tf.random_normal([self.mem_size, self.edim], stddev=self.init_std))

 self.T_B = tf.Variable(tf.random_normal([self.mem_size, self.edim], stddev=self.init_std))

 # m_i = sum A_ij * x_ij + T_A_i

 Ain_c = tf.nn.embedding_lookup(self.A, self.context)

 Ain_t = tf.nn.embedding_lookup(self.T_A, self.time)

 Ain = tf.add(Ain_c, Ain_t)

 # c_i = sum B_ij * u + T_B_i

 Bin_c = tf.nn.embedding_lookup(self.B, self.context)

 Bin_t = tf.nn.embedding_lookup(self.T_B, self.time)

 Bin = tf.add(Bin_c, Bin_t)

 for h in xrange(self.nhop):

 self.hid3dim = tf.reshape(self.hid[-1], [-1, 1, self.edim])

 Aout = tf.matmul(self.hid3dim, Ain, adjoint_b=True)

 Aout2dim = tf.reshape(Aout, [-1, self.mem_size])

 P = tf.nn.softmax(Aout2dim)

 probs3dim = tf.reshape(P, [-1, 1, self.mem_size])

 Bout = tf.matmul(probs3dim, Bin)

 Bout2dim = tf.reshape(Bout, [-1, self.edim])

 Cout = tf.matmul(self.hid[-1], self.C)

 Dout = tf.add(Cout, Bout2dim)

 self.share_list[0].append(Cout)

 if self.lindim == self.edim:

 self.hid.append(Dout)

 elif self.lindim == 0:

 self.hid.append(tf.nn.relu(Dout))

 else:

 F = tf.slice(Dout, [0, 0], [self.batch_size, self.lindim])

 G = tf.slice(Dout, [0, self.lindim], [self.batch_size, self.edim-self.lindim])

 K = tf.nn.relu(G)

 self.hid.append(tf.concat(axis=1, values=[F, K]))

3. Reference
Blog: Memory-network

Github: MemN2N-tensorflow

Paper: End-To-End Memory Networks

Deep learning 三巨头从机器学习谈起，指出传统机器学习的不足，总览深度学习理论、模型，给出了深度学习的发展历史，以及DL中最重要的算法和理论。
Supervised learning demo 监督学习案例 假设函数: 使用h(hypothesis, 假设)表示 输入(input value) 向量或者实数: 使用小写字母x等 矩阵: 使用大写字母X等
论文笔记系列-Neural Architecture Search With Reinforcement Learning 神经网络在多个领域都取得了不错的成绩，但是神经网络的合理设计却是比较困难的。在本篇论文中，作者使用 递归网络去省城神经网络的模型描述，并且使用 增强学习训练RNN，以使得生成得到的模型在验证集上取得最大的准确率。
论文笔记之：Collaborative Deep Reinforcement Learning for Joint Object Search Collaborative Deep Reinforcement Learning for Joint Object Search   CVPR 2017 Motivation： 传统的 bottom-up object region proposals 的方法，由于提取了较多的 proposal，导致后续计算必须依赖于抢的计算能力，如 GPU 等。