什么时候需要进行数据的标准化? 为什么？

TODO1/when-to-standardize-your-data.md

@@ -2,86 +2,86 @@
 > * 原文作者：[Zakaria Jaadi](https://medium.com/@zakaria.jaadi)
 > * 译文出自：[掘金翻译计划](https://github.com/xitu/gold-miner)
 > * 本文永久链接：[https://github.com/xitu/gold-miner/blob/master/TODO1/when-to-standardize-your-data.md](https://github.com/xitu/gold-miner/blob/master/TODO1/when-to-standardize-your-data.md)
-> * 译者：
+> * 译者：[Ultrasteve](https://github.com/Ultrasteve)
 > * 校对者：
 
-# When and why to standardize your data?
+# 什么时候需要进行数据的标准化? 为什么？
 
-> A simple guide on when to standardize your data and when not to.
+> 一份告诉你什么时候应该进行数据标准化的指南
 
 ![Credits : 365datascience.com](https://cdn-images-1.medium.com/max/NaN/1*dZlwWGNhFco5bmpfwYyLCQ.png)
 
-Standardization is an important technique that is generally performed as a pre-processing step before many Machine Learning models, to standardize the range of features of input data set.
+数据标准化是一种重要的技术，通常来说，在使用许多机器学习模型之前，我们都要使用它来对数据进行预处理, 它能对输入数据集里面的各个特征的范围进行标准化。
 
-Some ML developers tend to standardize their data blindly before “every” Machine Learning model without taking the effort to understand why it must be used, or even if it’s needed or not. So the goal of this post is to explain how, why and when to standardize data.
+一些机器学习工程师通常在使用所有机器学习模型之前，会盲目地对他们的数据进行标准化，然而，其实他们并不清楚数据标准化的理由，更不知道什么情况下使用这一技术是必要的，什么时候不是。因此，在这篇文章中，我们会阐述数据标准化的原因，及什么时候需要这样做，做法是什么？
 
-## Standardization
+## 标准化
 
-Standardization comes into picture when features of input data set have large differences between their ranges, or simply when they are measured in different measurement units (e.g., Pounds, Meters, Miles … etc).
+当输入数据的变化范围很大，或者它们各自使用的单位不同时（比如说一些用米，一些用厘米），我们会想到对数据进行标准化。
 
-These differences in the ranges of initial features causes trouble to many machine learning models. For example, for the models that are based on distance computation, if one of the features has a broad range of values, the distance will be governed by this particular feature.
+像这种初始特征变化范围较大的数据，会在我们使用许多机器学习模型时造成麻烦。例如，有一个模型是基于距离的，当其中一个特征值变化范围较大时，那么预测结果很大程度上就会受到它的影响。
 
-To illustrate this with an example: say we have a 2-dimensional data set with two features, Height in Meters and Weight in Pounds, that range respectively from [1 to 2] Meters and [10 to 200] Pounds. No matter what distance based model you perform on this data set, the Weight feature will dominate over the Height feature and will have more contribution to the distance computation, just because it has bigger values compared to the Height. So, to prevent this problem, transforming features to comparable scales using standardization is the solution.
+我们这里举一个例子。现在我们有一个二维的数据集，它有两个特征，以米为单位的高度（范围是 1 到 2 米）和以磅为单位的重量（范围是 10 到 200 磅）。不论你使用什么基于距离的模型，重量特征对结果的影响都会大大的高于高度特征，因为它的数据变化范围相对更大。因此，为了防止这种问题发生，我们会在这里用到数据标准化来约束重量特征的数据变化范围。
 
-## How to standardize data?
+## 如何进行数据标准化?
 
 ### Z-score
 
-Z-score is one of the most popular methods to standardize data, and can be done by subtracting the mean and dividing by the standard deviation for each value of each feature.
+`Z-score`是最受欢迎的数据标准化方法之一，在这种方法中，我们对每一项数据减去它的平均值并除以它的方差。
 
 ![](https://cdn-images-1.medium.com/max/NaN/0*AgmY9auxftS9BI73.png)
 
-Once the standardization is done, all the features will have a mean of zero, a standard deviation of one, and thus, the same scale.
+一旦完成了数据标准化, 所有特征对应的数据平均值变为 0 , 方差变为 1 , 因此, 所有特征的数据变化范围现在是一致的.
 
-> There exist other standardization methods but for the sake of simplicity, in this story i settle for Z-score method.
+> 其实还有许多数据标准化的方法，但为了降低难度，我们在这篇文章中只使用这种方法。
 
-## When to standardize data and why?
+## 什么时候需要进行数据的标准化? 为什么？
 
-As seen above, for distance based models, standardization is performed to prevent features with wider ranges from dominating the distance metric. But the reason we standardize data is not the same for all machine learning models, and differs from one model to another.
+就如上面所说，在基于距离的模型中，数据标准化用于防止大范围的特征对预测结果进行较大的影响。不过数据标准化的理由不仅仅只有这一个，对于不同的模型会有不同的理由。
 
-So before which ML models and methods you have to standardize your data and why?
+那么，在使用什么机器学习算法和模型之前，我们需要进行数据标准化呢？原因又是什么？
 
-**1- Before PCA:**
+**1- 主成分分析:**
 
-In Principal Component Analysis, features with high variances/wide ranges, get more weight than those with low variance, and consequently, they end up illegitimately dominating the First Principal Components (Components with maximum variance). I used the word “Illegitimately” here, because the reason these features have high variances compared to the other ones is just because they were measured in different scales.
+在主成分分析中，方差较大或者范围较大的特征，会相较于小方差小范围的数据获得更高的权重，这样会导致它们不合常理的主导第一主成分（方差最大的成分）的变化。为什么说这是不合常理的呢？因为导致这一特征比其他特征权重更大的理由，仅仅是因为它的数据变化范围更大。
 
-Standardization can prevent this, by giving same wheightage to all features.
+通过给予所有特征相同的权重，数据标准化可以防止这一点。
 
-**2- Before Clustering:**
+**2- 聚类:**
 
-Clustering models are distance based algorithms, in order to measure similarities between observations and form clusters they use a distance metric. So, features with high ranges will have a bigger influence on the clustering. Therefore, standardization is required before building a clustering model.
+聚类是一种基于距离的模型。为了测量观测对象之间的相似性，并将它们聚集在一起，模型需要测量它们之间的距离。在这种算法中，范围较大的特征会对聚类结果产生更大的影响。因此，在进行聚类之前我们需要进行数据标准化。
 
-**3- Before KNN:**
+**3- KNN:**
 
-k-nearest neighbors is a distance based classifier that classifies new observations based on similarity measures (e.g., distance metrics) with labeled observations of the training set. Standardization makes all variables to contribute equally to the similarity measures .
+k邻居算法是一种基于距离的分类算法，它通过测量新来的数据与已标记数据之间的距离来对其分类。数据标准化使得所有变量对测量结果产生同样的影响。
 
-**4- Before SVM**
+**4- SVM**
 
-Support Vector Machine tries to maximize the distance between the separating plane and the support vectors. If one feature has very large values, it will dominate over other features when calculating the distance. So Standardization gives all features the same influence on the distance metric.
+支持向量机尝试最大化决策平面与支持向量之间的距离。如果一个特征的值很大，那么相较于其他特征它会对计算结果造成更大的影响。数据标准化在这里使得所有特征对计算结果都能造成同样的影响。
 
 ![Credits : Arun Manglick ([arun-aiml.blogspot.com](http://arun-aiml.blogspot.com/))](https://cdn-images-1.medium.com/max/2000/0*_taflmQxrsa0vguT.PNG)
 
-**5- Before measuring variable importance in regression models**
+**5- 在回归模型中测量自变量的重要性**
 
-You can measure variable importance in regression analysis, by fitting a regression model using the **standardized** independent variables and comparing the absolute value of their standardized coefficients. But, if the independent variables are not standardized, comparing their coefficients becomes meaningless.
+你可以在回归分析中测量变量的重要程度。首先使用标准化过后的独立变量来训练模型，然后计算它们对应的标准化系数的绝对值差就能得出结论。然而，如果独立变量是未经标准化的，那比较它们的系数将毫无意义。
 
-**6- Before Lasso and Ridge Regression**
+**6- Lasso回归和岭回归**
 
-LASSO and Ridge regressions place a penalty on the magnitude of the coefficients associated to each variable. And the scale of variables will affect how much penalty will be applied on their coefficients. Because coefficients of variables with large variance are small and thus less penalized. Therefore, standardization is required before fitting both regressions.
+Lasso回归和岭回归对各变量对应的系数进行惩罚。变量的范围将会影响到他们对应系数受到什么程度的惩罚。因为方差大的变量对应的系数很小，因此它们会受到更小的惩罚。所以在进行以上算法之前进行数据标准化是必须的。
 
-## Cases when standardization is not needed?
+## 什么时候不需要标准化？
 
-**Logistic Regression and Tree based models**
+**逻辑回归和树形模型**
 
-Logistic Regression and Tree based algorithms such as Decision Tree, Random forest and gradient boosting, are not sensitive to the magnitude of variables. So standardization is not needed before fitting this kind of models.
+逻辑回归，树形模型（决策树，随机森林）和梯度提升对于变量的大小并不敏感。所以数据标准化在这里并不必要。
 
-## Conclusion
+## 结论
 
-As we saw in this post, when to standardize and when not to, depends on which model you want to use and what you want to do with it. So, it’s very important for a ML developer to understand the internal functioning of machine learning algorithms, to be able to know when to standardize data and to build a successful machine learning model.
+如上所说，正确使用数据标准化的时机取决于你当前在使用什么模型，你想用模型达到怎么样的目的。因此，如果机器学习工程师想知道什么时候该进行数据标准化并建造一个成功的机器学习模型，理解机器学习算法的内在原理十分重要。
 
-> N.B: The list of models and methods when standardization is required, presented in this post is not exhaustive.
+> 注: 这篇文章并没有列出所有需要标准化的模型和方法。
 
-### References:
+### 参考文献:
 
 * [**365DataScience.com**]: Explaining Standardization Step-By-Step
 * [**Listendata.com** ]: when and why to standardize a variable