Deep Learning Techniques for Text Classification

Evaluate the performance of TCN and Ensemble-based models using Word2Vec to your common deep learning architectures

Photo by Annie Spratt on Unsplash

Designed experiments to evaluate 20+ deep learning models on 5 text classification datasets in Python and TensorFlow.
Performed different text cleaning methods to fit the Bag-of-Words (BoW) and Word Embedding (WE) text representation.
Engineered features by using different BoW scoring methods and WE modes (random, static pre-trained, and dynamic pre-trained).
Optimized the feedforward, recurrent, convolutional, and ensemble-based neural networks to achieve the best model.
Attained the best average accuracy by 90% using an ensemble-based model and temporal convolutional network, competing with the state-of-the-art benchmark models.
Increased the average accuracy by 10% using a static pre-trained word embedding (Word2Vec), resulting in a 2x fold increase in performance from existing systems.

Recent Posts