Applied Data Analytics

COURSE  REGISTRATION OPEN
Course in English
Course description :

Introduction:

  • Course objectives
  • Structure of the course
  • AI, Machine Learning, Deep Learning
  • Supervised, Unsupervised Learning, Semi-supervised Learning
  • Classification and Regression
  • Generative Models
  • Examples of applications

Data Preparation, Data Formats, Feature Engineering:

  • Setting up a data analysis project
  • Project Goal Definition & Data Collection/Generation
  • Real data versus simulated data
  • Synthetic samples
  • Data representations
  • Data formats
  • Data preparation
  • Visualize data to gain insights
  • Handling of missing values
  • Data labelling
  • Manual feature engineering
  • Look at the distributions of features
  • Curse of dimensionality
  • Imbalanced Datasets
  • Ensemble Methods
  • Training, Validation and Test sets
  • Cross-validation

Classic Supervised Learning:

  • Use cases
  • Frame the problem
  • Prepare Data Vectors
  • Training Models:
    • Decision Trees, Random Forests
    • Ensemble Learning: Bagging, Boosting, Stacking
    • Artificial Neural Networks (ANNs)
  • Undefitting and Overfitting
  • Scikit-Learn and Keras
  • Fine-tune your model

Unsupervised Learning:

  • Unlabelled data & Unsupervised Learning
  • Data Inspection & Clustering
  • Using clustering for:
    • Preprocessing
    • Semi-supervised Learning
  • Anomaly detection

Convolutional Neural Networks (CNNs):

  • The Convolution operation
  • The Max pooling operation
  • Building your network
  • Data Augmentation
  • Visualizing what CNNs learn

Deep Learning for sequences:

  • Recurrent Neural networks
  • LSTM and GRU layers
  • Forecasting

Generative Models:

  • Feature Representation Learning
  • Density Estimation
SCHEDULE:

18-22 November 2024.  Details in this document

Email (contact pédagogique) : yvonne.becherini@apc.in2p3.fr

2 points

REGISTRATION: OPEN

For doctoral school STEPUP :  please register through the usual google sheet  with code:  ED-SPU30-STE30

For UPC students only , please also register through AMETHIS