Hello everyone,
I recently researched a Use Case on handwriting recognition using Principal Component Analysis (PCA) to extract image features and k-Nearest Neighbors ...
Hello everyone,
I recently researched a Use Case on handwriting recognition using Principal Component Analysis (PCA) to extract image features and k-Nearest Neighbors (kNN) for classification. This is a classic approach, but it explains the classification logic, unlike CNN models, which are often difficult to explain.
I would like to open a discussion on:
Share experiences when Granny Game applying PCA and kNN to handwriting or image data.
Discuss ways to reduce data dimensionality while maintaining high accuracy.
Compare the advantages and disadvantages between simple, explainable models and complex, high-accuracy models (like CNNs).
Everyone can contribute ideas on implementation methods, supporting tools, or sample data for testing. I believe that discussing this together will help everyone better understand the basic application of Machine Learning in real-world problems.
I look forward to receiving your feedback, sharing your experiences, and your experiments.