Scribd
Upload
11 views2 pages

Pneumonia Detection Using Convolutional Neural Networks

The document discusses the Elephant Herding Optimization (EHO) algorithm, which mimics the social behavior of elephant herds to solve optimization problems, particularly in pneumonia detection using Convolutional Neural Networks (CNN). Key components of EHO include clan updating and a separating operator for exploration, which enhance solution diversity and prevent local optima. The document compares traditional CNNs with Adam optimizer to CNN models optimized with EHO, highlighting the advantages of EHO in terms of exploration, convergence speed, and robustness.

Uploaded by

edukku22
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
11 views2 pages

Pneumonia Detection Using Convolutional Neural Networks

The document discusses the Elephant Herding Optimization (EHO) algorithm, which mimics the social behavior of elephant herds to solve optimization problems, particularly in pneumonia detection using Convolutional Neural Networks (CNN). Key components of EHO include clan updating and a separating operator for exploration, which enhance solution diversity and prevent local optima. The document compares traditional CNNs with Adam optimizer to CNN models optimized with EHO, highlighting the advantages of EHO in terms of exploration, convergence speed, and robustness.

Uploaded by

edukku22
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
You are on page 1/ 2

Pneumonia Detection Using Convolutional Neural Networks (CNN) Optimized with Elephant

Herding Optimization (EHO)-

Elephant Herding Optimization (EHO) is a nature-inspired algorithm based on how elephant herds
behave in real life. It’s a way to solve optimization problems by mimicking the social behavior of
elephants in their herds.

Key Terms and Variables

1. Number of Clans: C
The total number of herds or groups of elephants.

2. Elephants in Each Clan: n_e


The number of elephants in each clan.

3. Position of an Elephant: X_i


Each elephant represents a potential solution, and its position is represented by a vector of
numbers X_i.

4. Leader of a Clan (Matriarch): L_c


The best solution (highest fitness) in clan c.

5. Fitness Function: f(X)


A function that measures the quality of the elephant’s solution. Lower or higher values may
represent better solutions (depending on the problem).

Two Main Operators in EHO

o Clan Updating (Following the Leader)


Elephants in the same clan follow their leader (matriarch). The position of each
elephant in clan c is updated using the leader’s position

o β (beta) is a control parameter that decides how much the elephant should move
towards the leader.

o Random noise is added to encourage exploration.

Explanation:
Each elephant moves slightly closer to the leader with every iteration.

2. Separating Operator (Male Elephants Wandering)


Some male elephants (a small percentage) leave the herd and move to a random new
location:
Advantage:

1. Balanced Exploration and Exploitation: Prevents getting stuck in local optima while refining
the best solution.

2. Parallel Search: Multiple clans explore different regions, improving solution diversity.

3. Avoids Premature Convergence: Male elephants wandering ensures broader search space
exploration.

4. Simplicity: Easy to implement with straightforward update rules.

5. Flexibility: Can handle different types of optimization problems without much modification.

6. Adaptable Parameters: Tuning movement and separation rates allows better control over
behavior.

7. Faster Convergence: Clan behavior speeds up finding the best solution.

8. Robustness: Handles complex problem landscapes effectively.

9. Nature-Inspired Efficiency: Inspired by real-world elephant behavior, making it resilient in


dynamic environments.

Traditional CNN with Adam Optimizer vs CNN Model optimized with EHO:

Reference for Pneumonia Detection Using CNN based Feature Extraction-

https://www.semanticscholar.org/paper/Pneumonia-Detection-Using-CNN-based-Feature-Varshni-
Thakral/2e9bf21eaedf7a18e099d1e0d8faba07d36fc154

You might also like