Standard Scaler in SKLearn

• The sci kit learns in python is a library thatch is used in machine learning which is used to work on data modeling.
• It is only focused on the data modeling, it does not focus on manipulation and loading of the data.
• The classification, clustering through constancy interface, and regression are the modeling.

Uses of SK learn

• The main use of scikit learn in python is that it is open source.
• The benefit is:
  1. Open source
  2. Versatile used
  3. Free
  4. Easy to use
  5. Properly documented

Open source:

• The SK learn is a library that is open source and publicly available.
• Anyone can use this library and can modify the code.

Versatile used:

• The SK learn is a library that is used for many things like can identify user actions.
• Customer behaviour is also predicted which shows the versatile nature.
• SK learn library is user-friendly and comes with many tools.

Free:

• The SK learn is a library, which people use for free and any license for running is not necessary

Easy to use:

• The SK learn is a python library that people can access the library at any time.
• It is mostly used by research organizations for their operation since SK learn is easy to use.

Properly documented:

• The SK learn is a python library in which documentation is used properly.
• It contains API documentation which is accessible from the website provided.

Advantages of SK learn

• It is user-friendly and can do many things like can identify user actions.
• It is free to use.
• It is easy to use.
• It provides API documentation for the user.
• Here the documentation is done properly.
• The contributor updated the SK learn library done by the international community.
• It is spread for the BSD license which means it can used  without restriction.

Disadvantages of SK learn

• It is not much of use for in-depth learning.

Let us consider an example to see the working of the SK learn library:

from sklearn.cluster import kmeans_plusplus
from sklearn.datasets import make_blobs
import matplotlib.pyplot as pl
/ / here sample means it is used to generate sample data
sample = 5000
component = 4
/ / kmeans_plusplus  which is used to seed calculate from k
a, b_true = make_blobs( n_samples = sample,centers = component, cluster_std = 0.60, random_state = 0 )
a = a[ : , : : -1]
centers_init, indices = kmeans_plusplus(a, n_clusters = 4, random_state = 0)
/ / plot figure means it is used for seed plotting on the screen
plot.figure(1)
colors =["red", "blue", "green", "yellow"]
 
for k, col in enumerate(colors):
cluster_data = b_true == k
plot.scatter( a [cluster_data, 0], a [cluster_data, 1], c=col, marker=".", s=10)
 
plot.scatter(centers_init[:, 0], centers_init[:, 1], c="b", s=50)
/ / plot tile is used to give the title for the graph on the screen
plot.title("K-Mean Clustering")
plot.xticks([])
plot.yticks([])
plot.show()

output:

Installing the SK learn:

• SK learn is used to model the data.To install the SK learn library we need to install scipy and numpy.
• Command to install the numpy is:
  pip install numpy
• After execution of the above command, we can see the numpy is installed.
• After numpy is installed we need to install scipy. We can install scipy using the following command.
  pip install scipy
• After execution of the above command, the scipy is installed. After both libraries are installed. We can install scikit library using the following command:
  pip install scikit-learn
• If the SK learn is already installed then no need to install them again.
• We can update the SK learn using the following command:
  pip install -u scikit-learn
  SK learn features:
• We know that SK learn is used to model the data but not to manipulate the data.
• The following are the features of SK learn:
  Supervised learning:
• It is a predictive model, in which data comes with good quality which we want to predict.
• It is divided into two types:
  1. Classification:
     • if the problem output has been categorized into “white”, and”black” then the problem is considered to be a classification problem.it is a predictive model in which the set data is categorized into classes.
  2. Regression:
     • If the problem output is continuous output then the problem is considered to be a regression problem. Examples are “kilometers”.

   Dimension reduction:

• In mathematics the dimension has measured the distance and size of the object.
• We can reduce the number of input features from the set using dimension reduction which is predictive modelling.

Cross-validation:

• The cross-validation is used to make sure that the model is accurate since predictive modeling is done by supervised learning.
• It is used to predict data of the predictive modeling.
  Open source:
• The program code is freely available or anyone can access it since it is open source.
• The user can not only access but also can modify the code since is made for public use.

Feature selection:

• The SK learn is used to select the features from the set.
• The predictor object that is used by the made is used to tunning down.
• The supervise model is used to recognize the attribute.

Ensemble methods:

• The machine learning technique is used to combine many models which is turned intothe predictive model, this model is called the ensemble model.
• In other words we call consider it as the combination of prediction of models.
  Clustering:
• The problem in which we want to find the inherent is called the clustering problem.
• Grouping the customers by their behaviour which is permanent.
  Unsupervised learning:
• It does not provide any guidance and does not provide any superior. here the labels are not present for data.
• For a variable x there is no output respectively since there is no superior.
• Here the model is needed to allowed to work on own data.