← Back to selected work03 / MACHINE LEARNING

Whirling Triangle-Based Pattern Classification

Exploring rotational geometric features and machine-learning classifiers for pattern classification.

ROTATIONAL FEATURES / CLASSIFICATION
BEST RESULT66.39% accuracy
CLASSIFIERSVM · RBF kernel
METHODRotational feature extraction
OVERVIEW

Geometry meets machine learning.

Developed a triangle-transformation-based pattern classification project using Python and machine learning. The workflow explored rotational feature extraction and classification, with a Support Vector Machine using the RBF kernel achieving the highest reported accuracy of 66.39%.

APPROACH
01Triangle transformation

Generate transformed geometric patterns for classification.

02Rotational feature extraction

Extract features based on rotational transformations.

03ML classification

Apply machine-learning classification to the extracted features.

04Evaluation

Compare model performance; SVM (RBF) gave the best reported result.

STACK
PythonMachine LearningSVMRBF Kernel
Next project — Gesture Control System ↗