Skip to content

Latest commit

 

History

History
27 lines (22 loc) · 488 Bytes

File metadata and controls

27 lines (22 loc) · 488 Bytes

Algorithms

  • Feature Descriptor
    1. HOG
    2. LBP
  • Classification
    1. Decision Tree
    2. Gradient Boosting
    3. K-Nearest Neighbor
    4. Random Forest

Techniques

  1. Machine Learning
  2. Computer Vision
  3. GridSearchCV
  4. Image Augmentation

Previous Results

  1. Gradient Boosting got
    • accuracy → 86.67%
    • sensitivity → 93.33%
    • specificity → 80.33%
  2. KNN got
    • accuracy → 89.33%
    • sensitivity → 91.67%