articulus_divisio

Tandem Approaches

Classification of multiple text datasets using various algorithms, including

• Kmeans
• Agglomerative Clustering
  • Ward
  • Complete
  • Single
  • Average
• HDBSCAN
• Spectral Clustering
• Gaussian Mixtures

The classification is based on multiple work representations including:

• Word2Vec
• GloVe
• BERT
• ROBERTA

Represented in various spaces by using dimensionality reduction techniques including:

• PCA
• t-SNE
• UMAP
• Simple Autoencoder

First Submission:

Labeled Data (Classic4 and BBC)

Second Submission

Labeled Data (Classic4 and BBC)

Unlabeled Data

Articles1

Articles2

Simultaneous Dimensionality Reduction and Classification

Classification of multiple text datasets using various algorithms, including

• Reduced k-means et Factorial k-means
• Deep Clustering Network (DCN)
• Deep k-means (DKM)

The classification is based on multiple work representations including:

• Word2Vec
• GloVe
• BERT
• ROBERTA

First Submission:

Labeled Data (Classic4 and BBC)

Second Submission:

Labeled Data (Classic4 and BBC)

Unlabeled Data (Articles1 and Articles2)