MBD4
DATA ANALYSIS AND MACHINE LEARNING

<Training programs

Objectives :

Acquisition of the basics of machine learning

Level of trainees :

Dataminers, statisticians, business intelligence consultants.

MBD1, MSG2 training and basic knowledge in statistics are mandatory as prerequisite.

Duration and location :

• Duration : 21 hours over 3 days
• Location: Soladis training center in Paris or Lyon (to consult the calendar, click on “Sign up”) or directly in your office (minimum of 4 people)

Training fee :

• In Soladis office : 2400€ tax-free per trainee
• Other : Contact us for an offer adapted to your company.

Soladis invites you to participate in this training course on Big Data analysis and machine learning methods, which aim is to give you a cross-sectional overview of the methods adapted to data valuation. This training addresses an informed public wishing to understand and deepen its knowledge on the subject.

Training Schedule

INTRODUCTION

• Why Data Science and Machine Learning are important today ?
• History: from the beginnings of Artificial Intelligence to Deep Learning
• The Data Scientist’s job: tasks and tools
• Descriptive statistics recall and synthetic representation of data (functions, distribution, estimation, correlation/regression)

MACHINE LEARNING ALGORITHMS: THEORY AND APPLICATION IN R

• Supervised learning: forecasting and classification (linear regression, logistic regression, Support Vector Machine, KNN, Decision Trees and Random Forests)
• Unsupervised learning: clustering and dimensionality reduction (KMeans, Hierarchical Clustering, SVM and PCA, Manifold Learning)
• Deep Learning and Neural Networks
• Reinforcement learning
• Text mining and Hidden Markov Models

DIFFERENT KIND OF VISUALIZATION

• Classic Visualization
• Modern visualization
• Interactive visualization

WHICH ALGORITHM IS THE BEST ?

• The overfitting risk and how to get around it
• Define a train/test approach
• Hyperparameters tuning

THE CONSTRAINTS AND LIMITS OF MACHINE LEARNING

A FIRST MACHINE LEARNING PROJECT

• An example of machine learning application through a Kaggle project

Do not hesitate to contact us for more information about this training, which can be organized in our training center (dates provided on our website) or directly in your premises, according to your preferences.