IT Expertise - Machine Learning

General Data
Academic program	ECAM ENGINEERING PROGRAM			:
Type d'EC	Classes
Lectures : 8h00 Lab Work : 12h00 Total duration : 26h00	Status :	Period : SEMESTER 8	Education language : English

Learning Outcomes
* Understand and acquire knowledge of basic concepts in Machine Learning • Build hypothesis based on a context • Test models and evaluate the consistency of their results with the application objectives • Understand the stakes and risks of artificial decision making * Be able to analyse and solve ML related problems (text classification, image classification, recommendation, regression...), using for instance Scikit Learn, Keras and TensorFlow library or NLP specific tools.

Learning Outcomes

* Understand and acquire knowledge of basic concepts in Machine Learning
• Build hypothesis based on a context
• Test models and evaluate the consistency of their results with the application objectives
• Understand the stakes and risks of artificial decision making

* Be able to analyse and solve ML related problems (text classification, image classification, recommendation, regression...), using for instance Scikit Learn, Keras and TensorFlow library or NLP specific tools.

Content
* Introduction to IA and Machine Learning ° Model Based Learning - main concepts and definitions ° Exact solution, iterative solutions, Gradient Descent ° First algorithms: Regressions - Linear Regression - Logistic Regression ° Data preprocessing ° Hyperparameter tuning * Towards Deep Learning ° more on preprocessing (categories encoding) and data sets ° From biological neuron to perceptron ° Multilayer Perceptron ° Convolutional Neural Networks ° Transfer Learning * Introduction to Natural Language Processing ° Some important ideas about NLP ° Example of statistical NLP (Multinomial Naive Bayes) ° NLP with Deep Learning (LSTM) * Other algorithms * Group project (text classification, image classification, recommendation, regression...)

Content

* Introduction to IA and Machine Learning
° Model Based Learning - main concepts and definitions
° Exact solution, iterative solutions, Gradient Descent
° First algorithms: Regressions
- Linear Regression
- Logistic Regression
° Data preprocessing
° Hyperparameter tuning

* Towards Deep Learning
° more on preprocessing (categories encoding) and data sets
° From biological neuron to perceptron
° Multilayer Perceptron
° Convolutional Neural Networks
° Transfer Learning

* Introduction to Natural Language Processing
° Some important ideas about NLP
° Example of statistical NLP (Multinomial Naive Bayes)
° NLP with Deep Learning (LSTM)

* Other algorithms

* Group project (text classification, image classification, recommendation, regression...)

Pre-requisites / co-requisites
* Matrix algebra and calculus * Probability and Statistics * Algorithmic and programming

Bibliography
the 100 page machine Learning Book, Andriy Burkov Deep Learning, Ian Goodfellow and Yoshua Bengio and Aaron Courville (2016 - MIT press) Hands-on Machine Learning with Scikit-Learn, Keras and TensorFlow, Aurélien Geron (2019 - O Reilly) Deep Learning With Python, François Chollet, (2021, Manning)

Assessment(s)
N°	Nature	Coefficient	Observable objectives
1	Group project by teams of 3, to analyse some given dataset. The enthusiasm in experimenting different models, finding ways to enhance first results will be mostly appreciated. Part of the presentation should be dedicated to knowledge/concepts/ideas that were discovered during the project.	100	Project
2	At the end of each session, students have to send a report of their lab work.	1	Continuous Assessment
3	Continuous Assessment	1