nuluh/thesis
1
0
Fork 0
Code Issues 47 Pull Requests Actions Packages Projects Releases 1 Wiki Activity

Enhance Thesis with Normalization References #12

New Issue
Open
opened 2024-08-24 17:54:24 +00:00 by nuluh · 0 comments
nuluh commented 2024-08-24 17:54:24 +00:00 (Migrated from github.com)
Copy Link

#11

Problem Statement

The current thesis lacks a detailed discussion on normalization techniques, which are crucial for improving the performance and accuracy of Support Vector Machines (SVM). Specifically, when dealing with vibration data in damage localization prediction, the variations in feature scale can significantly impact the model's ability to learn effectively. Incorporating references to normalization methods will not only strengthen the theoretical foundation but also align with best practices in the field.

Context from Research

According to Chapter 2, Section 2.2 of Hsu et al.'s "A Practical Guide to Support Vector Classification," scaling (often referred to as normalization) plays a critical role in SVM performance. The authors emphasize that feature scaling ensures all input features contribute equally to the decision function. Without scaling, features with larger numeric ranges could dominate the learning process, leading to suboptimal decision boundaries.

The section highlights two main types of scaling:

  1. Min-Max Scaling: Transforms features to a fixed range, often [0, 1], which preserves the relationships between feature values.
  2. Z-Score Normalization: Adjusts features based on their mean and standard deviation, resulting in a mean of 0 and a standard deviation of 1.

The choice of scaling method can depend on the nature of the features and the specific problem domain.

Suggested Writings

In the revised thesis, include the following text in the section discussing feature engineering:

### Feature Scaling in Damage Localization Prediction

In the context of damage localization using SVM, it is essential to preprocess the input features through scaling. As highlighted by Hsu et al. (2003), scaling helps in mitigating the dominance of features with larger numeric ranges, thereby allowing the SVM to establish more balanced and effective decision boundaries.

For this thesis, two scaling techniques are considered:

- **Min-Max Scaling**: This method scales the input features to a fixed range, typically [0, 1]. It is particularly useful when the input features have a known range, and the relationships between feature values need to be preserved.

- **Z-Score Normalization**: This approach scales the features based on their mean and standard deviation, ensuring a normalized distribution with a mean of 0 and a standard deviation of 1. This technique is effective when dealing with features that may have different units or varying statistical distributions.

Given the nature of vibration data in structural health monitoring, where the magnitude of sensor readings can vary significantly, Z-Score Normalization is particularly suitable. It allows the SVM model to treat each feature with equal importance, improving the overall accuracy of damage localization predictions.

This addition will demonstrate an understanding of the importance of feature scaling in SVM and reinforce the technical depth of the thesis.

#11 ## Problem Statement The current thesis lacks a detailed discussion on normalization techniques, which are crucial for improving the performance and accuracy of Support Vector Machines (SVM). Specifically, when dealing with vibration data in damage localization prediction, the variations in feature scale can significantly impact the model's ability to learn effectively. Incorporating references to normalization methods will not only strengthen the theoretical foundation but also align with best practices in the field. ## Context from Research According to Chapter 2, Section 2.2 of Hsu et al.'s ["A Practical Guide to Support Vector Classification,"](https://www.semanticscholar.org/paper/A-Practical-Guide-to-Support-Vector-Classification-Hsu-Chang/8b9cd90af0631bbed04c0718230f0faed1eca209) **scaling** (often referred to as normalization) plays a critical role in SVM performance. The authors emphasize that feature scaling ensures all input features contribute equally to the decision function. Without scaling, features with larger numeric ranges could dominate the learning process, leading to suboptimal decision boundaries. The section highlights two main types of scaling: 1. **Min-Max Scaling**: Transforms features to a fixed range, often [0, 1], which preserves the relationships between feature values. 2. **Z-Score Normalization**: Adjusts features based on their mean and standard deviation, resulting in a mean of 0 and a standard deviation of 1. The choice of scaling method can depend on the nature of the features and the specific problem domain. ## Suggested Writings In the revised thesis, include the following text in the section discussing feature engineering: ```markdown ### Feature Scaling in Damage Localization Prediction In the context of damage localization using SVM, it is essential to preprocess the input features through scaling. As highlighted by Hsu et al. (2003), scaling helps in mitigating the dominance of features with larger numeric ranges, thereby allowing the SVM to establish more balanced and effective decision boundaries. For this thesis, two scaling techniques are considered: - **Min-Max Scaling**: This method scales the input features to a fixed range, typically [0, 1]. It is particularly useful when the input features have a known range, and the relationships between feature values need to be preserved. - **Z-Score Normalization**: This approach scales the features based on their mean and standard deviation, ensuring a normalized distribution with a mean of 0 and a standard deviation of 1. This technique is effective when dealing with features that may have different units or varying statistical distributions. Given the nature of vibration data in structural health monitoring, where the magnitude of sensor readings can vary significantly, Z-Score Normalization is particularly suitable. It allows the SVM model to treat each feature with equal importance, improving the overall accuracy of damage localization predictions. ``` This addition will demonstrate an understanding of the importance of feature scaling in SVM and reinforce the technical depth of the thesis.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
Something isn't working
 critical
Block Progress
 documentation
Improvements or additions to documentation
 documentclass
Improvements or additions to LaTeX documentclass
 duplicate
This issue or pull request already exists
 enhancement
New feature or request
 good first issue
Good for newcomers
 help wanted
Extra attention is needed
 invalid
This doesn't seem right
 latex
Related to the writings
 question
Further information is requested
 wontfix
This will not be worked on
No Label enhancement latex
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
nuluh
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: nuluh/thesis#12
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?