Accès direct au contenu


Version anglaise


Accueil > Formations > Master MVA > Présentation des cours

Advanced learning for text and graph data ALTEGRAD

Lecturer : Michalis Vazirgiannis (Polytechnique)

Introduction :

The ALTEGRAD course ( 28 hours) aims at providing an overview of state-of-the-art ML and AI methods for text and graph data with a significant focus on applications. Each session will comprise two hours of lecture followed by two hours of programming sessions.

Grading for the course will be based on a final data challenge plus lab based evaluation.


All interested students should fill the following form to enrol: link to form

The ALTEGRAD challenge Fall 2017 topic is:
"Can you predict whether two short texts have the same meaning?"

You can find more information for the challenge here

Course web page: here

Course Syllabus 2019-20 :

1.1 Graph-of-words

        - Graph-of-words  and information retrieval

        -Keyword extraction, Summarization (abstrative, extractive)
-Graph based document categorization

-Event detection in text streams (twitter)

1.2  Deep Learning for Text Mining and NLP
  1.2.1. Word & document embeddings

-Introduction to NN architectures for NLP

-Latent Semantic Indexing

-Word and document embeddings embeddings: word2vec, glove models, word mover's distance, doc2vec, subword based embeddings, context based embeddings (ELMO)
           1.2.2. Document similarity metrics: word mover's distance, etc.

           1.2.3 Deep Learning architectures for text classification (CNNs, RNNs, LSTMs) and text generations (sequence to sequence, attention networks..)

2.Deep Learning for Graphs

2.1 Graph kernels

    - Graph Generators,   Graph Similarity with graph kernels, community detection, kernel similarity frameworks, Grakel library

2.2 Deep Learning for Graphs (DeepWalk, Node Embeddings
    - DeepWalk, LINE, GraRep, node2vec, Graph Classification with  Patchy-San, Spectral Networks 
    - node embeddings (deepwalk & node2vec) for node classification and link prediction
    - Supervised node embeddings (GCNN, ...)

    - Graph classification, Kernel Graph NNs, message passing

    - Graph autoencoders and applications

2.3 Deep learning for Influence maximization