You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Built a CNN that classifies arabic handwritten characters that achieved a 94% accuracy on the test set after applying transfer learning and data augmentation.
Arabic sentiment classifier on real-world company reviews β TextVectorization + Embedding + LSTM, fine-tuned with Keras Tuner RandomSearch, achieving 86% test accuracy.