Scribd
Upload
15 views

Support_Vector_Regression_Introduction

Uploaded by

sathyanarayana.thirunagari
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
15 views

Support_Vector_Regression_Introduction

Uploaded by

sathyanarayana.thirunagari
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 10

Introduction to Support Vector

Regression (SVR)
A Machine Learning Approach for
Regression Analysis
[Your Name/Institution]
What is Support Vector Regression
(SVR)?
• SVR is a regression algorithm that
approximates relationships between input
variables and continuous target variables.
• Goal: Minimize prediction error while fitting
data points in continuous space.
SVR vs. SVM
• Support Vector Machines (SVM) - Used for
classification tasks, finds hyperplane to
separate classes.
• Support Vector Regression (SVR) - Used for
regression, fits hyperplane within tolerance
margin for continuous outcomes.
• Key Difference: SVM for discrete classes, SVR
for continuous predictions.
How SVR Works
• Hyperplane: Represents the best fit in high-
dimensional space.
• Decision Boundary: Lines at distance ‘a’
(epsilon) from hyperplane, containing most
data points.
• Objective: Fit hyperplane with minimal error
within boundaries.
SVR Mathematical Model
• Hyperplane Equation: Y = wx + b
• Decision Boundary Equations:
• - wx + b = +a
• - wx + b = -a
• Error Constraint: Ensures most points satisfy -a
< Y - (wx + b) < +a
Implementing SVR in Python
• 1. Step 1: Import Libraries - NumPy,
Matplotlib, Pandas
• 2. Step 2: Load Dataset - Position_Salaries.csv
• 3. Step 3: Feature Scaling
• 4. Step 4: Fit SVR Model - Using ‘rbf’ kernel
• 5. Step 5: Predict New Result - Predict salary
for position level 6.5
• 6. Step 6: Visualize Results - Plot SVR fit with
plt.plot
Code Snippets for Python
Implementation
• • Importing Libraries
• • Loading Data
• • Feature Scaling
• • Fitting SVR Model
• • Prediction and Inverse Transform
• • Visualization
Key Concepts in SVR
• Kernel Function: Allows for non-linear
relationships (e.g., radial basis function,
linear).
• Epsilon: Margin of tolerance around the
hyperplane.
• Support Vectors: Points closest to the
hyperplane influencing its position.
Key Takeaways
• SVR’s Versatility: Handles non-linear data,
suitable for complex relationships.
• Hyperparameters: Kernel choice and epsilon
crucial for model performance.
• Applications: Used in finance, engineering,
healthcare for predictive analytics.
Conclusion
• SVR extends SVM principles for continuous
predictions with various kernels.
• Strengths: Flexible, robust, accurate for
regression tasks.
• Applications: Ideal for complex datasets across
multiple fields.

You might also like