Unsupervised learning is a machine learning technique in which the dataset has no target variable or no response value\(Y \).The data is unlabelled. Simply saying,there is no target value to supervise the learning process of a learner unlike in supervised learning where we have training examples which have both input variables \(X_i\) and target variable\(Y\) i,e\( {(xi,yi)} \) vectors and by looking and learning from the training examples the learner generates a mapping function(also called a hypothesis) \( f:X_i−>Y \) which maps \(X_i\) values to \(Y\) and learns the relationship between input variables and target variable so that we could generalize it to some random unseen test examples and predict the target value.
The best example of unsupervised learning is when a small child is given some unlabelled pictures of cats and dogs , so only by looking at the structural and visual similarities and dissimilarities between the images , the child classifies one as a dog and other as cat.
Unsupervised learning is inclined towards finding groups and subgroups from data by finding the associations, similarities and relationships between the inputs \(X_i\). It is important for understanding the variations and grouping structure of a dataset and is also used as a preprocessing tool for finding the best and most important features \(X_i\) which explain the most variance and summarize the most information in the data using techniques such as principal component analysis(PCA) for supervised learning techniques.
exampleIf we have a dataset with 100 predictors and we wanted to generate a model,it would be highly inefficient to use all those 100 predictors because that would increase the variance and complexity of the model and which in turn would lead to overfitting. Instead, what PCA does is find 10 most correlated variables and linearly combine them to generate principal components \(Z_m\) which could be further used as features for our model.
Principal Component Analysis
PCA introduces a lowerdimensional representation of the dataset.It finds a sequence of linear combination of the variables called the principal components\(Z_1,Z_2…Z_m\) that explain the maximum variance and summarize the most information in the data and are mutually uncorrelated.
What we try to do is find most relevant set of variables and simply linearly combine the set of variables into a single variable\(Z_m\) called a principal component.

1) The first principal component \(PC_1\) has the highest variance across data.
2) The second principal component \(PC_2\) is uncorrelated with \(PC_1\) which also has high variance.
We have tons of correlated variables in a high dimensional dataset and what PCA tries to do is pair and combine them to a set of some important variables that summarize all information in the data.
PCA will give us new set of variables called principal components which could further be used as inputs in a supervised learning model. So now we have lesser and most important set of variables paired together to form a new single variable which explains most variance in data. This technique is often termed as dimensionality reduction which is famous technique to do feature selection/reduction and use only relevant features \(X_i\) in the model.
Details
We have a set of input vectors \( x_1,x_2,x_3…..x_p\) with \(n\) observations in dataset.
The 1st principal component \(Z_1\) of a set of features is the normalized linear combination of the features \(x_1,x_2….x_p\).
$$Z_1 = \sum_{i=1}^p \phi_{1}x_1 + \phi_{{2}}x_2 + \phi_{3}x_3 + ………\phi_{i}x_i $$
where n=no of observations, p = number of variables. It is a linear combination to find out the highest variance across data. By normalized it means \( \sum_{j=1}^{p} \phi_{j}^2 = 1\).
We refer to the weights \( \phi_{n X p}\) as Loading matrix.The loadings make up the principal components loading vector. \( \phi_1 = ( \phi_{11},\phi_{21},\phi_{31} , \phi_{41}……,\phi_{n1} )^T \) is the loading vector for \(PC_1\).
We constrain the loadings so that their sum of squares could be 1, as otherwise setting these elements to be arbitrarily large in absolute value could result in an arbitrarily large variance
The first Principal component solves the below optimization problem of maximizing variance across the components
$$maximize: \frac{1}{n} \sum_{i=1}^n \sum_{j=1}^p (\phi_{ji}.X_{ij})^2 subject \ to \sum_{j=1}^p \phi_{ji}^2=1 $$
Here each principal component has mean 0.
The above problem can be solved via Single value decomposition of matrix \(X\) ,which is a standard technique in linear algebra.
Enough maths now let’s start implementing PCA in R.
Implementing PCA in R
We will use USAarrests data.
?USArrests #dataset which contains Violent Crime Rates by US State dim(USArrests) dimnames(USArrests) ## [1] 50 4 ## [[1]] ## [1] "Alabama" "Alaska" "Arizona" "Arkansas" ## [5] "California" "Colorado" "Connecticut" "Delaware" ## [9] "Florida" "Georgia" "Hawaii" "Idaho" ## [13] "Illinois" "Indiana" "Iowa" "Kansas" ## [17] "Kentucky" "Louisiana" "Maine" "Maryland" ## [21] "Massachusetts" "Michigan" "Minnesota" "Mississippi" ## [25] "Missouri" "Montana" "Nebraska" "Nevada" ## [29] "New Hampshire" "New Jersey" "New Mexico" "New York" ## [33] "North Carolina" "North Dakota" "Ohio" "Oklahoma" ## [37] "Oregon" "Pennsylvania" "Rhode Island" "South Carolina" ## [41] "South Dakota" "Tennessee" "Texas" "Utah" ## [45] "Vermont" "Virginia" "Washington" "West Virginia" ## [49] "Wisconsin" "Wyoming" ## ## [[2]] ## [1] "Murder" "Assault" "UrbanPop" "Rape"
Finding means for all variables.
#finding mean of all apply(USArrests,2,mean) ## Murder Assault UrbanPop Rape ## 7.788 170.760 65.540 21.232
Finding variance of all variables.
apply(USArrests,2,var) ## Murder Assault UrbanPop Rape ## 18.97047 6945.16571 209.51878 87.72916
There is a lot of difference in variances of each variables. In PCA mean does not play a major role, but variance plays a major role in defining principal components so very large differences in variance value of a variable will definately dominate the principal components. We need to standardize the variables so as to get mean \(\mu=0\) and variance \(\sigma^2=1\). To standardize we use formula \(x’ = \frac{x – mean(x)}{sd(x)}\).
The function prcomp()
will do the needful of standardizing the variables.
pca.out<prcomp(USArrests,scale=TRUE) pca.out ## Standard deviations (1, .., p=4): ## [1] 1.5748783 0.9948694 0.5971291 0.4164494 ## ## Rotation (n x k) = (4 x 4): ## PC1 PC2 PC3 PC4 ## Murder 0.5358995 0.4181809 0.3412327 0.64922780 ## Assault 0.5831836 0.1879856 0.2681484 0.74340748 ## UrbanPop 0.2781909 0.8728062 0.3780158 0.13387773 ## Rape 0.5434321 0.1673186 0.8177779 0.08902432
#summary of the PCA summary(pca.out) ## Importance of components%s: ## PC1 PC2 PC3 PC4 ## Standard deviation 1.5749 0.9949 0.59713 0.41645 ## Proportion of Variance 0.6201 0.2474 0.08914 0.04336 ## Cumulative Proportion 0.6201 0.8675 0.95664 1.00000
names(pca.out) ## [1] "sdev" "rotation" "center" "scale" "x"
Now as we can see maximum % of variance is explained by \(PC_1\), and all PCs are mutually uncorrelated. Around 62 % of variance is explained by \(PC_1\).
Let’s build a biplot to understand better.
biplot(pca.out,scale = 0, cex=0.65)
Now in the above plot red colored arrows represent the variables and each direction represent the direction which explains the most variation. Example for all the countries in the direction of ‘UrbanPop’ are countries with most urbanpopulation and opposite to that direction are the countries with least. So this is how we interpret our Biplot.
Conclusion
PCA is a great preprocessing tool for picking out the most relevant linear combination of variables and use them in our predictive model.It helps us find out the variables which explain the most variation in the data and only use them. PCA plays a major role in the data analysis process before going for advanced analytics and model building.
The only drawback PCA has is that it generates the principal components in a unsupervised manner i.e without looking the target vector \( (y_1,y_2,y_3…..y_n) \) ,hence the principal components which explains the most variation in dataset without looking at the target\(Y\) variable,may or may not explain good percentage of variance for the response variable \(Y\) which could affect and degrade the performance of the predictive model.
The R code for implementing PCA in R is adapted from the amazing online course “Statistical learning” offered by Stanford University Online. I urge readers to definately go and try out this course to get clear with the core statistics and maths behind various statistical models. The details about the course can be found hereStatistical Learning
Also, this book Elements Of statistical learning has helped me learn lots of amazing stuff
Hope you guys liked the article, make sure to like and share it. Happy machine learning!!
Related Post
 Find Your Best Customers with Customer Segmentation in Python
 ARIMA models and Intervention Analysis
 A Gentle Introduction on Market Basket Analysis — Association Rules
 Building A Book Recommender System – The Basics, kNN and Matrix Factorization
 Explaining complex machine learning models with LIME