Keywords: machine learning, deep-learning, patient stratification, data reduction and correlation

Methods:

  • Data reduction and correlation
    • PCA, PLS, t-SNE, LASSO,..; 
    • Canonical correlation;
  • Unsupervised clustering:
    • Bayesian modeling;
    • Gaussian Mixture models;
    • Dirichlet processes;
    • Graph-based methods, Community detection; 
    • Outlier detection;
  • Supervised learning:
    • Random Forests, SVM, ...;
    • Survival models;
    • Regression;
  • Deep-learning
    • Feature extraction, segmentation, classification and regression;
    • Adversarial models and GAN;
    • Activation maps, uncertainty maps,...

     

    Programming languages:

    Python, R, Pytorch, Keras

    Points of contact:

    • Dr Elsa Angelini
    • Dr Sam Greenbury
    • Dr Paul Blakeley