Machine Learning Techniques
In this notebook we perform some of the techniques that helps us to make our workflow more efficient. We starts out with the simple one and move on with some advance techniques related to Hyperparameter Tuning or Encoding. We divide this notebook into sub-section with the heading at the top followed by the description and some sort of sample code.
Agenda:
- Use of Lambda Function for performing single line operation
- Understand the
pandas.DataFrame.groupby
aggregate function - Learn how to perform Target Encoding
- Learn the Optimization Technique other than RandomizedSearchCV and GridSearchCV
- Learn how to fill missing values using the Machine Learning Model
0. Use of Lambda Function for performing single line operation
In this section, we perform some mathematical operation on the dataset using the Lambda Function
. We Perform the operation, create new columns using the map
and the apply
function.
Lets prepare some data over which we are going to perform the Lambda Function operation. For this we are using the Sklearn Dataset library and import the california housing
dataset.
# import the library
import pandas as pd
import numpy as np
from sklearn.datasets import fetch_california_housing
import matplotlib.pyplot as plt
%matplotlib inline
dataset = fetch_california_housing()
df = pd.DataFrame(dataset.data, columns=dataset.feature_names)
df['MedHouseVal'] = dataset.target
df.head()
We prepare our dataset on which we gonna work with. Now we perfrom:
- Apply the
min-max normalization
onpopulation
column. - Multiply the
AveRooms
with 100. - Create the bins for the
Housege
column.