{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import pandas as pd#导入csv文件的库\nimport math\nimport numpy as np#进行矩阵运算的库\nimport matplotlib.pyplot as plt#作图的库\nimport torch#一个深度学习的库Pytorch\nimport torch.nn as nn#neural network,神经网络\nimport torch.optim as optim#一个实现了各种优化算法的库\nimport torch.nn.functional as F#神经网络函数库\nimport warnings#避免一些可以忽略的报错\nwarnings.filterwarnings('ignore')","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2023-06-27T12:12:43.743973Z","iopub.execute_input":"2023-06-27T12:12:43.744376Z","iopub.status.idle":"2023-06-27T12:12:43.750475Z","shell.execute_reply.started":"2023-06-27T12:12:43.744345Z","shell.execute_reply":"2023-06-27T12:12:43.749386Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#设置随机种子\nimport random\ntorch.backends.cudnn.deterministic = True#将cudnn框架中的随机数生成器设为确定性模式\ntorch.backends.cudnn.benchmark = False#关闭CuDNN框架的自动寻找最优卷积算法的功能，以避免不同的算法对结果产生影响\ntorch.manual_seed(42)\nnp.random.seed(42)\nrandom.seed(42)","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:43.752632Z","iopub.execute_input":"2023-06-27T12:12:43.753447Z","iopub.status.idle":"2023-06-27T12:12:43.767601Z","shell.execute_reply.started":"2023-06-27T12:12:43.753279Z","shell.execute_reply":"2023-06-27T12:12:43.766659Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train=pd.read_csv(\"/kaggle/input/ece597697-sp2023/train.csv/train.csv\")\ntrain.head()","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:43.769585Z","iopub.execute_input":"2023-06-27T12:12:43.770035Z","iopub.status.idle":"2023-06-27T12:12:50.198999Z","shell.execute_reply.started":"2023-06-27T12:12:43.769964Z","shell.execute_reply":"2023-06-27T12:12:50.197959Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"datay=train['label'].values\ntrain=train.drop(['label','id'],axis=1)\ndataX=train.values","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.232475Z","iopub.execute_input":"2023-06-27T12:12:50.232925Z","iopub.status.idle":"2023-06-27T12:12:50.358824Z","shell.execute_reply.started":"2023-06-27T12:12:50.232891Z","shell.execute_reply":"2023-06-27T12:12:50.357551Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#划分训练集和测试集的函数\ndef train_test_split(dataX,datay,shuffle=True,percentage=0.8):\n    \"\"\"\n    将训练数据X和标签y以numpy.array数组的形式传入\n    划分的比例定为训练集:测试集=8:2\n    \"\"\"\n    if shuffle:\n        random_num=[index for index in range(len(dataX))]\n        np.random.shuffle(random_num)\n        dataX=dataX[random_num]\n        datay=datay[random_num]\n    split_num=int(len(dataX)*percentage)\n    train_X=dataX[:split_num]\n    train_y=datay[:split_num]\n    test_X=dataX[split_num:]\n    test_y=datay[split_num:]\n    return train_X,train_y,test_X,test_y","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.360482Z","iopub.execute_input":"2023-06-27T12:12:50.361296Z","iopub.status.idle":"2023-06-27T12:12:50.368754Z","shell.execute_reply.started":"2023-06-27T12:12:50.361249Z","shell.execute_reply":"2023-06-27T12:12:50.367500Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_X,train_y,valid_X,valid_y=train_test_split(dataX,datay,percentage=0.9)\nlabel_count=np.max(train_y)+1\nfor index in range(label_count):\n    print(f\"np.sum(train_y==index)={np.sum(train_y==index)}\")\nprint(f\"train_X.shape:{train_X.shape},valid_X.shape:{valid_X.shape}\")","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.370819Z","iopub.execute_input":"2023-06-27T12:12:50.371322Z","iopub.status.idle":"2023-06-27T12:12:50.380494Z","shell.execute_reply.started":"2023-06-27T12:12:50.371211Z","shell.execute_reply":"2023-06-27T12:12:50.379159Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#多头注意力机制\nclass MultiHeadSelfAttention(nn.Module):\n    #定义初始化函数\n    def __init__(self,dim_in,d_model,num_heads=4):\n\n        super(MultiHeadSelfAttention,self).__init__()\n        \n        self.dim_in=dim_in\n        \n        self.d_model=d_model\n        \n        self.num_heads=num_heads\n        \n        #向量的维度必须被头的个数整除,否则会抛出异常.\n        assert d_model %num_heads==0,\"d_model must be multiple of num_heads\"\n        \n        #定义线性变换矩阵\n        self.linear_q=nn.Linear(dim_in,d_model)\n        self.linear_k=nn.Linear(dim_in,d_model)\n        self.linear_v=nn.Linear(dim_in,d_model)\n        self.scale=1/math.sqrt(d_model//num_heads)\n        \n        #最后的线性层\n        self.fc=nn.Linear(d_model,d_model)\n        \n    def forward(self,x):\n        batch,n,dim_in=x.shape\n        \n        assert dim_in==self.dim_in\n        \n        nh=self.num_heads\n        \n        dk=self.d_model//nh\n        \n        q=self.linear_q(x).reshape(batch,n,nh,dk).transpose(1,2)\n        k=self.linear_k(x).reshape(batch,n,nh,dk).transpose(1,2)\n        v=self.linear_v(x).reshape(batch,n,nh,dk).transpose(1,2)\n        \n        \n        dist=torch.matmul(q,k.transpose(2,3))*self.scale\n        dist=torch.softmax(dist,dim=-1)\n        \n        att=torch.matmul(dist,v)\n        \n        att=att.transpose(1,2).reshape(batch,n,self.d_model)\n        \n        output=self.fc(att)\n        \n        return output ","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.397089Z","iopub.execute_input":"2023-06-27T12:12:50.397506Z","iopub.status.idle":"2023-06-27T12:12:50.412363Z","shell.execute_reply.started":"2023-06-27T12:12:50.397473Z","shell.execute_reply":"2023-06-27T12:12:50.411374Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class SELayer(nn.Module):\n    def __init__(self,channel=256,reduction=16):\n        super(SELayer,self).__init__()\n        self.avg_pool=nn.AdaptiveAvgPool2d(1)#输出为1*1\n        self.fc=nn.Sequential(\n            nn.Linear(channel,channel//reduction,bias=True),\n            nn.BatchNorm1d(channel//reduction),\n            nn.GELU(),\n            nn.Linear(channel//reduction,channel,bias=True),\n            nn.Sigmoid(),\n        )\n    def forward(self,x):\n        #batch_size,channels,\n        b,c,_,_=x.shape\n        \n        y=self.avg_pool(x).view(b,c)\n        y=self.fc(y).view(b,c,1,1)\n        return x*y.expand_as(x)#将y扩展到x同样的shape","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.416250Z","iopub.execute_input":"2023-06-27T12:12:50.417062Z","iopub.status.idle":"2023-06-27T12:12:50.427501Z","shell.execute_reply.started":"2023-06-27T12:12:50.417030Z","shell.execute_reply":"2023-06-27T12:12:50.426833Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#对特征图的每个位置也提取一次权重\nclass SEConv(nn.Module):\n    def __init__(self,area=36,reduction=4):\n        super(SEConv,self).__init__()\n        self.fc=nn.Sequential(\n            nn.Linear(area,area//reduction,bias=True),\n            nn.BatchNorm1d(area//reduction),\n            nn.GELU(),\n            nn.Linear(area//reduction,area,bias=True),\n            nn.Sigmoid(),\n        )\n    def forward(self,x):\n        #batch_size,channels,(height*width)\n        b,c,area=x.shape\n        y=torch.mean(x,dim=1)\n        y=self.fc(y)\n        y=y.view(-1,1,area)\n        return x*y.expand_as(x)#将y扩展到x同样的shape","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.429093Z","iopub.execute_input":"2023-06-27T12:12:50.429800Z","iopub.status.idle":"2023-06-27T12:12:50.444198Z","shell.execute_reply.started":"2023-06-27T12:12:50.429760Z","shell.execute_reply":"2023-06-27T12:12:50.443065Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class Active_Rotating_Filters(nn.Module):\n    #定义初始化函数\n    def __init__(self,in_channels=3,out_channels=64,kernel_size=3,stride=1,padding=0):\n        super(Active_Rotating_Filters,self).__init__()\n        self.in_channels=in_channels\n        self.out_channels=out_channels\n        self.kernel_size=kernel_size\n        self.stride=stride\n        self.padding=padding\n        \n        self.conv=nn.Conv2d(in_channels,out_channels,kernel_size,stride,padding)\n        self.rot_conv=nn.Conv2d(in_channels,out_channels,kernel_size,stride,padding)#用于存放旋转后卷积的参数\n    \n    #对卷积核的权重进行旋转\n    def rotate_weight(self,index,weight):#out_channel*in_channel*kernel_size*kernel_size\n        #逆时针旋转90度旋转一次\n        rot_weight=torch.rot90(weight,index,(2,3))\n        return rot_weight\n        \n    def forward(self,x):\n        rot_x=self.conv(x)\n        for index in range(1,4):\n            #旋转后的参数\n            self.rot_conv.weight=nn.Parameter(self.rotate_weight(index,self.conv.weight))\n            rot_x=torch.cat((rot_x,self.rot_conv(x)),dim=1)\n        b,c,h,w=rot_x.shape\n        rot_x=rot_x.reshape(b,c//4,4,h,w)\n        rot_x,_=torch.max(rot_x,dim=2)#,_忽略索引\n        rot_x=rot_x.view(b,-1,h,w)\n        return rot_x","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#score:0.92976\nclass CNN(nn.Module):\n\n    def __init__(self):\n        \n        super(CNN,self).__init__()\n        \n        self.flag=0\n        self.center_array=[]\n        \n        #创建自注意机制的类\n        self.Self_Attention=MultiHeadSelfAttention(dim_in=36,d_model=36,num_heads=3)\n        \n        self.selayer=SELayer()\n        \n        self.seconv=SEConv()\n        \n        self.conv=nn.Sequential(\n        Active_Rotating_Filters(1,16,3,1,0),#30*30->28*28\n        nn.BatchNorm2d(16),\n        nn.SiLU(),\n        Active_Rotating_Filters(16,32,3,1,0),#28*28->26*26\n        nn.BatchNorm2d(32),\n        nn.ReLU(),\n        Active_Rotating_Filters(32,72,5,1,0),#26*26->22*22\n        nn.BatchNorm2d(72),\n        nn.Tanh(),\n        nn.Conv2d(72,128,5,1,1),#20*20\n        nn.GELU(),\n        nn.MaxPool2d(2,2,0),#10*10\n        nn.BatchNorm2d(128),\n        nn.Conv2d(128,256,5,1,0),#6*6\n        nn.SiLU(),\n        nn.BatchNorm2d(256),\n        self.selayer,\n        )\n        \n        self.fc=nn.Sequential( \n        nn.Linear(256*6*6,128),\n        nn.BatchNorm1d(128),\n        nn.Dropout(p=0.2),\n        nn.GELU(),\n        nn.Linear(128,10),\n        )\n    def circle_padding(sellf,image, pad=1):\n        \"\"\"\n        input: torch.Tensor image (batch_size, channels, height, width)\n        output: pad_image (batch_size, channels, height+2*pad, width+2*pad)\n        \"\"\"\n        b, c, h, w = image.shape\n        pad_image = torch.zeros((b, c, h+2*pad, w+2*pad), device=image.device)\n\n        pad_image[:, :, pad:h+pad, pad:w+pad] = image\n        pad_image[:, :, 0:pad, pad:w+pad] = image[:, :, -pad:, :]\n        pad_image[:, :, -pad:, pad:w+pad] = image[:, :, 0:pad, :]\n        pad_image[:, :, pad:h+pad, 0:pad] = image[:, :, :, -pad:]\n        pad_image[:, :, pad:h+pad, -pad:] = image[:, :, :, 0:pad]\n        pad_image[:, :, 0:pad, 0:pad] = image[:, :, -pad:, -pad:]\n        pad_image[:, :, -pad:, -pad:] = image[:, :, 0:pad, 0:pad]\n        pad_image[:, :, 0:pad, -pad:] = image[:, :, -pad:, 0:pad]\n        pad_image[:, :, -pad:, 0:pad] = image[:, :, 0:pad, -pad:]\n\n        return pad_image\n    \n    def get_centering_array(self,image):\n        if self.flag==0:\n            #传入的是np.array的数组\n            batch_size,channels,row,col=image.shape\n            center_x=row/2\n            center_y=col/2\n            array=torch.randn(row,col).to(device)\n            for i in range(row):\n                for j in range(col):\n                    array[i][j]=math.sqrt(1- ( (i-center_x)**2+(j-center_y)**2)/( center_x**2+center_y**2 )/10)\n            self.flag=1\n            self.center_array=array\n        return image*self.center_array\n    #噪音注入\n    def noise_injection(self,features,noise_factor=0.002):\n        noise = torch.randn_like(features) * noise_factor\n        return features + noise\n    \n    def forward(self,input): \n        output=self.circle_padding(input)\n        output=self.get_centering_array(output)\n        output=self.noise_injection(output)\n        output=self.conv(output)\n        output=output.view(-1,256,6*6)#reshape\n        output=self.seconv(output)\n        output=output+self.Self_Attention(output)#跳连接\n        output=output.view(-1,256*6*6)#reshape\n        output=self.fc(output)\n        return F.log_softmax(output/5,dim=1)","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.445869Z","iopub.execute_input":"2023-06-27T12:12:50.446597Z","iopub.status.idle":"2023-06-27T12:12:50.462554Z","shell.execute_reply.started":"2023-06-27T12:12:50.446558Z","shell.execute_reply":"2023-06-27T12:12:50.461740Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%%time\n\ndevice='cuda' if torch.cuda.is_available() else\"cpu\"\n\n\ntrain_X=np.array(train_X).reshape(-1,1,28,28)/255.0-0.5\ntrain_y=np.array(train_y)\nvalid_X=np.array(valid_X).reshape(-1,1,28,28)/255.0-0.5\n\nvalid_X1=torch.Tensor(valid_X).to(device)\nnetC=CNN()\ntrain_accs=[]#存储训练集的准确率\nvalid_accs=[]#存储测试集的准确率\n#训练周期为3000次\nnum_epochs=3000\nbatch_size=200#一次训练的数据量\nweight_decay=0.001\n\n#优化器\noptimizer=optim.Adam(netC.parameters(),lr=0.0005,betas=(0.5,0.999))\n#损失函数\ncriterion=nn.CrossEntropyLoss()\nnetC=netC.to(device)\nfor epoch in range(num_epochs):\n    \n    random_num=[index for index in range(len(train_X))]\n    np.random.shuffle(random_num)\n    \n    train_X=train_X[random_num]\n    \n    train_y=train_y[random_num]\n    \n    train_X1=torch.Tensor(train_X[:batch_size]).to(device)\n    train_y1=torch.Tensor(train_y[:batch_size]).long().to(device)\n    \n    #训练\n    netC.train()\n    #将数据放进去训练\n    output=netC(train_X1).to(device)\n    #计算每次的损失函数\n    loss=criterion(output,F.one_hot(train_y1,10).type(torch.FloatTensor).to(device)).to(device)\n    #L2正则化\n    l2_regularization = torch.tensor(0.).to(device)\n    for param in netC.parameters():\n        l2_regularization += torch.norm(param, 2)\n    loss += weight_decay * l2_regularization\n    #反向传播\n    loss.backward()\n    #优化器进行优化(梯度下降,降低误差)\n    optimizer.step()\n    #将梯度清空\n    optimizer.zero_grad()\n    if epoch%50==0:\n        netC.eval()\n        with torch.no_grad():\n            output=netC(train_X1).to(device)\n            output=output.detach().cpu().numpy()\n            train_pred=np.argmax(output,axis=1)\n\n            output=netC(valid_X1).to(device)\n            output=output.detach().cpu().numpy()\n            valid_pred=np.argmax(output,axis=1)\n        \n        train_acc=np.sum(train_pred==train_y[:batch_size])/len(train_y[:batch_size])\n        valid_acc=np.sum(valid_pred==valid_y)/len(valid_y)\n\n        train_accs.append(train_acc)\n        valid_accs.append(valid_acc)\n        print(f\"epoch:{epoch},loss:{loss}\\n train_acc:{train_acc},valid_acc:{valid_acc}\")","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:12:50.464100Z","iopub.execute_input":"2023-06-27T12:12:50.465186Z","iopub.status.idle":"2023-06-27T12:17:32.706272Z","shell.execute_reply.started":"2023-06-27T12:12:50.465151Z","shell.execute_reply":"2023-06-27T12:17:32.705230Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"epochs=[i*50 for i in range(len(train_accs))]\nplt.title(\"Fashion_mnist_dataset\")\nplt.plot(epochs,train_accs,marker=\"o\",label=\"train_accs\")\nplt.plot(epochs,valid_accs,marker=\"x\",label=\"valid_accs\")\nplt.legend()\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:17:32.707621Z","iopub.execute_input":"2023-06-27T12:17:32.708477Z","iopub.status.idle":"2023-06-27T12:17:33.052007Z","shell.execute_reply.started":"2023-06-27T12:17:32.708441Z","shell.execute_reply":"2023-06-27T12:17:33.050940Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#根据数据的标签和预测结果绘制混淆矩阵\ndef confusion_matrix(label,prediction):\n    \"\"\"\n    label:也就是每个数据真实的标签类别,np.array的一维数组\n    prediction:针对每个数据的预测结果,np.array的一维数组\n    \"\"\"\n    #简单做个判断,预测的个数和标签的个数要一致.\n    assert len(label)==len(prediction),\"The length of prediction and label must be the same.\"\n    num_classes=len(np.unique(label))#根据真实的标签看看有几类\n    matrix=np.zeros((num_classes,num_classes))#构建num_classes*num_classes混淆矩阵.\n    #真实标签放x轴,预测结果放y轴\n    for i in range(len(label)):\n        matrix[prediction[i]][label[i]]+=1\n    #log(x+1)\n    for i in range(num_classes):\n        for j in range(num_classes):\n            matrix[i][j]=np.log(matrix[i][j]+1)\n    #隐藏坐标轴的数字\n    #fig是画布,ax是坐标轴\n    fig,ax=plt.subplots()\n    #设置x轴和y轴的刻度\n    ax.set_xticks([])\n    ax.set_yticks([])\n    #设置颜色的调色板\n    plt.imshow(matrix,cmap='Blues')\n    #添加颜色条到图片\n    plt.colorbar()\n    #给混淆矩阵添加数字\n    for i in range(num_classes):\n        for j in range(num_classes):\n            plt.text(i, j, str(int(np.exp(matrix[j, i])-1)), ha='center', va='center', color='red')\n    #设置标签\n    plt.xlabel('True')\n    plt.ylabel('Predicted')","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:17:33.053741Z","iopub.execute_input":"2023-06-27T12:17:33.054428Z","iopub.status.idle":"2023-06-27T12:17:33.065512Z","shell.execute_reply.started":"2023-06-27T12:17:33.054391Z","shell.execute_reply":"2023-06-27T12:17:33.064499Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"netC.eval()\nwith torch.no_grad():\n    output=netC(valid_X1).to(device)\n    output=output.detach().cpu().numpy()\n    valid_pred=np.argmax(output,axis=1)","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:17:33.068238Z","iopub.execute_input":"2023-06-27T12:17:33.068955Z","iopub.status.idle":"2023-06-27T12:17:33.958888Z","shell.execute_reply.started":"2023-06-27T12:17:33.068917Z","shell.execute_reply":"2023-06-27T12:17:33.957744Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"confusion_matrix(valid_y,valid_pred)","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:17:33.960198Z","iopub.execute_input":"2023-06-27T12:17:33.960778Z","iopub.status.idle":"2023-06-27T12:17:34.667918Z","shell.execute_reply.started":"2023-06-27T12:17:33.960735Z","shell.execute_reply":"2023-06-27T12:17:34.666945Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test=pd.read_csv(\"/kaggle/input/ece597697-sp2023/test.csv/test.csv\")\ntest=test.drop(['id'],axis=1)\ntest_X=test.values/255.0-0.5\ny_pred=[]\nfor i in range(len(test_X)):\n    netC.eval()\n    with torch.no_grad():\n        image=test_X[i].reshape((1,1,28,28))\n        pred=netC(torch.Tensor(image).to(device))[0]\n        pred=np.argmax(pred.detach().cpu().numpy())\n        y_pred.append(pred)","metadata":{"execution":{"iopub.status.busy":"2023-06-27T12:17:34.669518Z","iopub.execute_input":"2023-06-27T12:17:34.669898Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#写入文件\nimport csv\nhead=[\"id\",\"label\"]\ndata=[]\nfor index in range(len(y_pred)):\n    data.append([index,y_pred[index]])\nwith open(\"fashion_mnsit_submission.csv\",'w',newline='') as f:\n    writer=csv.writer(f)\n    writer.writerow(head)\n    writer.writerows(data)","metadata":{"trusted":true},"execution_count":null,"outputs":[]}]}