How to deal with imbalance classes with downsampling in Python?
DATA MUNGING DATA CLEANING PYTHON MACHINE LEARNING RECIPES PANDAS CHEATSHEET     ALL TAGS

How to deal with imbalance classes with downsampling in Python?

How to deal with imbalance classes with downsampling in Python?

This recipe helps you deal with imbalance classes with downsampling in Python

0

Recipe Objective

While working on classification problem have you ever come across a bias dataset which contains most samples of a particular class. So to transform the dataset such that it contains equal number of classes in target value we can downsample the dataset. Downsampling means to reduce the number of samples having the bias class.

This data science python source code does the following:
1. Imports necessary libraries and iris data from sklearn dataset
2. Use of "where" function for data handling
3. Downsamples the higher class to balance the data

So this is the recipe on how we can deal with imbalance classes with downsampling in Python.

Step 1 - Import the library

import numpy as np from sklearn import datasets

We have imported numpy and datasets modules.

Step 2 - Setting up the Data

We have imported inbuilt wine datset form the datasets module and stored the data in x and target in y. This dataset is not bias so we are making it bias for better understanding of the functions, we have removed first 30 rows by selecting the rows after the 30 rows. Then in the selected data we have changed the class which are not 0 to 1. wine = datasets.load_wine() X = wine.data y = wine.target X = X[30:,:] y = y[30:] y = np.where((y == 0), 0, 1) print("Viewing the imbalanced target vector:\n", y)

Step 3 - Downsampling the dataset

First we are selecting the rows where target values are 0 and 1 in two different objects and then printing the number of observations in the two objects. w_class0 = np.where(y == 0)[0] w_class1 = np.where(y == 1)[0] n_class0 = len(w_class0) n_class1 = len(w_class1) print("n_class0: ", n_class0) print("n_class1: ", n_class1) In the output we will see the number of samples having target values as 1 are much more greater than 0. So in downsampling we will randomly select the number of rows having target as 1 and make it equal to the number of rows having taregt values 0.
Then we have printed the joint dataset having target class as 0 and 1. w_class1_downsampled = np.random.choice(w_class1, size=n_class0, replace=False) print(); print(np.hstack((y[w_class0], y[w_class1_downsampled]))) So the output comes as:

Viewing the imbalanced target vector:
 [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1]

n_class0:  29

n_class1:  119

[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1]

Relevant Projects

Customer Churn Prediction Analysis using Ensemble Techniques
In this machine learning churn project, we implement a churn prediction model in python using ensemble techniques.

Identifying Product Bundles from Sales Data Using R Language
In this data science project in R, we are going to talk about subjective segmentation which is a clustering technique to find out product bundles in sales data.

German Credit Dataset Analysis to Classify Loan Applications
In this data science project, you will work with German credit dataset using classification techniques like Decision Tree, Neural Networks etc to classify loan applications using R.

Machine Learning or Predictive Models in IoT - Energy Prediction Use Case
In this machine learning and IoT project, we are going to test out the experimental data using various predictive models and train the models and break the energy usage.

Demand prediction of driver availability using multistep time series analysis
In this supervised learning machine learning project, you will predict the availability of a driver in a specific area by using multi step time series analysis.

Time Series Forecasting with LSTM Neural Network Python
Deep Learning Project- Learn to apply deep learning paradigm to forecast univariate time series data.

Census Income Data Set Project - Predict Adult Census Income
Use the Adult Income dataset to predict whether income exceeds 50K yr based on census data.

Predict Macro Economic Trends using Kaggle Financial Dataset
In this machine learning project, you will uncover the predictive value in an uncertain world by using various artificial intelligence, machine learning, advanced regression and feature transformation techniques.

Ensemble Machine Learning Project - All State Insurance Claims Severity Prediction
In this ensemble machine learning project, we will predict what kind of claims an insurance company will get. This is implemented in python using ensemble machine learning algorithms.

Build an Image Classifier for Plant Species Identification
In this machine learning project, we will use binary leaf images and extracted features, including shape, margin, and texture to accurately identify plant species using different benchmark classification techniques.