qq_43118572的博客

1. Github代码# -*- coding: utf-8 -*-"""Created on Mon Sep 28 10:28:06 2020@author: wb"""import torchimport torch.nn as nnfrom GCN_models import GCNfrom One_hot_encoder import One_hot_encoderclass SSelfAttention(nn.Module): def __init__(self,

1. 对稀疏矩阵的正则化- 对称邻接矩阵的正则化# [A * D^(-1/2)]^T * D^(-1/2) = D^(-1/2) * A * D^(-1/2)def sym_adj(adj): """Symmetrically normalize adjacency matrix.""" # 压缩的邻接矩阵 adj = sp.coo_matrix(adj) rowsum = np.array(adj.sum(1)) # 将n*1的矩阵转换为1个向量

1. 数据的准备DCRNN的github项目有完整的数据Graph Wavenet提供的数据缺少 '/data/sensor_graph/adj_mx.pkl'

序列模型常常会用到padding（nn.functional.pad()），而padding添加的就是0.0。例如： def train(self, input, real_val): self.model.train() self.optimizer.zero_grad() input = nn.functional.pad(input, (1,0,0,0)) output = self.model(input)

from tensorflow.keras.layers import Dense, Lambda, Dot, Activation, Concatenatefrom tensorflow.keras.layers import Layerclass Attention(Layer): def __init__(self, units=128, **kwargs): self.units = units super().__init__(**kwargs)

RMSERMSE = Sqrt(1/N * sigma(vi - v^l)^ 2)MAEMAE = 1/N sigma|vi - v^l|MAPEMAPE = 1/N sigma[ |vi - v^l| / vi ]

SVM的图例学习from sklearn.datasets._samples_generator import make_blobsimport matplotlib.pyplot as pltimport numpy as npX, y = make_blobs(n_samples=200, centers=2, random_state=0, cluster_std=0.80)plt.scatter(X[:,0], X[:,1], c=y, s=50, cmap="autumn")plt

main.pyimport numpy as np# pickle用于进行序列化与反序列化# 序列化过程将文本信息转变为二进制数据流。这样就信息就容易存储在硬盘之中，# 当需要读取文件的时候，从硬盘中读取数据，然后再将其反序列化便可以得到原始的数据。import pickleimport osimport treePlotter# 创建训练数据def CreateTrainingDataset(): X = [[0, 2, 0, 0, 'N'], [0, 2,

Downstream taskDownstream tasks is what the field calls those supervised-learning tasks that utilize a pre-trained model or component简而言之，下游任务就是利用预训练的模型在当前数据集的效果Downstream task in CVDownstream Task: Downstream tasks are computer vision applications

VAEimport tensorflow as tfclass VariationalAutoencoder(object): def __init__(self, n_input, n_hidden, optimizer = tf.train.AdamOptimizer()): self.n_input = n_input self.n_hidden = n_hidden network_weights = self._initialize_w