matlab科研常用语法1

在数据分析之前，我们通常需要先将数据标准化，利用标准化后的数据进行数据分析；

```matlab
A = [1,2,3; 4,5,6;7,8,9];
tA = zscore(A)
```

```matlab
tA =

    -1    -1    -1
     0     0     0
     1     1     1
```

```
PS:个人感觉相差不是很大，有时候准确率可能会降低，也可能会好一些
```

matlab k近邻

[neighbor,distence] = knnsearch(kdtree,train_data,'k',k+1);
neighbor = neighbor(:,2:k+1);
distence = distence(:,2:k+1);

均值

mean(distence(:, k));

利用 k近邻构建相似度矩阵

[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-WmFPdtd7-1658234727324)(…/AppData/Roaming/Typora/typora-user-images/image-20220719155958072.png)]

for i = 1:p_data_num
    for j = 1:k
        S(i,neighbor(i,j)) = exp(-distence(i,j)*distence(i,j)/(sigma*sigma));
    end
end

标签和示例数

[label_num,p_data_num] = size(train_p_target);

label_num,p_data_num] = size(train_p_target);