Scribd
Upload
4 views8 pages

Experiment 5

The document outlines a lab activity focused on implementing and comparing techniques for handling imbalanced datasets in machine learning. It discusses various methods such as Random Under Sampling, Random Over Sampling, SMOTE, and Class Weighting, along with their implications on model performance. The conclusion emphasizes the importance of using appropriate evaluation metrics and highlights that SMOTE and Class Weighting yield better results for imbalanced data.

Uploaded by

shivangraturi15
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
4 views8 pages

Experiment 5

The document outlines a lab activity focused on implementing and comparing techniques for handling imbalanced datasets in machine learning. It discusses various methods such as Random Under Sampling, Random Over Sampling, SMOTE, and Class Weighting, along with their implications on model performance. The conclusion emphasizes the importance of using appropriate evaluation metrics and highlights that SMOTE and Class Weighting yield better results for imbalanced data.

Uploaded by

shivangraturi15
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

SUBJECT NAME – Elements of AI - ML

LAB ACTIVITY – 5

Implementation of Different Techniques for


Handling Imbalanced Data

SUBMITTED BY:
NAME – SHIVANG RATURI
SAP ID – 590022331
BATCH – 46

SUBMITTED TO:
DR. SANDEEP CHAND KUMAIN

1. Aim
To implement and compare different techniques for
handling imbalanced datasets using machine learning.

2. Theory
In machine learning, an imbalanced dataset occurs
when one class has significantly more samples than the
other.
Example:
 Class 0 → 90%
 Class 1 → 10%
This causes models to become biased toward the
majority class.
Techniques to Handle Imbalance
1. Random Under Sampling
2. Random Over Sampling
3. SMOTE (Synthetic Data Generation)
4. Class Weighting

3. Requirements
pip install numpy pandas matplotlib seaborn scikit-learn
imbalanced-learn

4. Implementation
Step 1: Import Libraries
Step 2: Create Imbalanced Dataset

Step 3: Visualize Data

Step 4: Train-Test Split


[Link] Training

6. Techniques Implementation

A. Random Under Sampling

B. Random Over Sampling


C. SMOTE

D. Class Weighting

[Link] of Results
8. Observations
 Original model shows high accuracy but poor
minority prediction
 Under Sampling loses data
 Over Sampling may cause overfitting
 SMOTE improves generalization
 Class Weighting balances learning effectively

9. Conclusion
 Handling imbalanced data is crucial in ML
 SMOTE and Class Weighting give better
performance
 F1-score is a better metric than accuracy for
imbalanced data

OUTPUT:
LEARNING OUTCOMES

 Understand imbalanced datasets and their impact on machine


learning model performance.
 Use appropriate evaluation metrics like precision, recall, F1-
score, and ROC-AUC instead of accuracy.
 Apply techniques to handle imbalance, such as oversampling,
undersampling, and class weighting.
 Compare and interpret model performance to select the most
effective method for a given problem.

You might also like