# 数据集扩充

`horizontally flipping` `random crops` `color jittering`

# 预处理(normalization)

` 减去均值` `zscore` `白化(whittening)`

## 减去均值

` >> x -= np..mean(X, axis = 0) # zero-center`

## zscore

```>> X -= np.mean(X, axis = 0) # zero-center
>> X /= np.std(X, axis = 0) # normalize```

## 白化(whitening)

1. `PCA白化`
2. `ZCA白化`

### PCA whitening

pca白化是指对上面的pca的新坐标X’,每一维的特征做一个标准差归一化处理

### ZCA whitening

ZCA白化是在PCA白化的基础上，又进行处理的一个操作。具体的实现是把上面PCA白化的结果，又变换到原来坐标系下的坐标:

## 权重的初始化

Small Random Numbers

`>> 0.01 * N(0,1)    #N(0,1)表示均值为0的标准高斯分布`

Calibrating the Variances

`方差(variance)为2/n`

`>> w = np.random.randn(n) * sqrt(2.0/n) # current recommendation`

# Training

## Filter size

it is important to employ a small filter (e.g., `3*3`) and small strides (e.g., 1) with zeros-padding, which not only reduces the number of parameters, but improves the accuracy rates of the whole deep network. Meanwhile, a special case mentioned above, i.e., `3*3` filters with stride 1, could preserve the spatial size of images/feature maps. For the pooling layers, the common used pooling size is of `2*2`.

## Activation Functions

• `sigmoid`很少用,不推荐(kill gradients, not zero-centered)
• `tanh``sigmoid`要好(is zero-centered)
• `ReLU`系列: `ReLU`,`PReLU`,`Leaky ReLU`,`RReLU`,中推荐使用`PReLU` and `RReLU`

l2正则一般情况下优于l1正则

0.5的概率值是典型的做法

## 数据倾斜

### sampling techniques

1. duplicating instances(maybe special crops processing) from the minority classes until a balanced distribution is reached (oversampling) 2. removing instances from over-represented classes (undersampling) 3. it is suggested that a combination is the best solution for extremely imbalanced distributions 4. generating new data in minority classes based on the current data

### cost sensitive techniques

a higher penalty can be given to the network when it misclassifies the minority classes during training

### One-class learning

1. only provides training data from a single class(每一类去训练,不停的fine-tuning)
2. firstly fine-tune on the classes which have a large number of training samples (images/crops), and secondly, continue to fine-tune but on the classes with limited number samples

89 篇文章41 人订阅

0 条评论

## 相关文章

1443

2765

1929

### 统计学习方法概论

1.统计学习 统计学习的对象是数据，它从数据出发，提取数据的特征，抽象出数据的模型，发现数据中的知识，又回到对数据的分析与预测中去。统计学习...

2734

### Single Shot MultiBox Detector论文翻译——中文版

SSD: Single Shot MultiBox Detector 摘要 我们提出了一种使用单个深度神经网络来检测图像中的目标的方法。我们的方法命名为SSD，...

2500

21410

751

### Single Shot MultiBox Detector论文翻译——中英文对照

SSD: Single Shot MultiBox Detector Abstract We present a method for detecting ob...

2120