Data science
Getting started with Data Science
From Python to your first machine learning model: a hands-on path into data.
- Duration
- 35 h · 5 days
- Level
- Beginner
- Format
- On site · Remote · Blended
- Max. group size
- 12 participants
Overview
This program introduces data science from scratch. Participants discover the lifecycle of a data project, learn the basics of Python and object-oriented programming, manipulate and visualize data with pandas and Matplotlib, acquire the essential statistics concepts, then build and evaluate their first machine learning models with scikit-learn. Each concept is immediately applied to real datasets, up to a complete mini-project presented at the end of the program.
Who this program is for
- Employees moving into or retraining for data careers
- Analysts, students and professionals curious about data science
- Training centers offering an introductory data path
Prerequisites
- No programming knowledge required
- High-school level math basics (percentages, averages, simple functions)
- Comfortable with a spreadsheet
Learning objectives
- Describe data careers and the lifecycle of a data project
- Write simple Python programs with Jupyter
- Structure your Python code with object-oriented programming (classes, inheritance, encapsulation)
- Clean, transform and analyze data with pandas
- Produce clear visualizations with Matplotlib and Seaborn
- Apply descriptive statistics and basic notions of inference
- Build, evaluate and interpret a machine learning model
- Carry out an end-to-end data mini-project
Detailed curriculum
10 modules · 35 h · 5 days
Day 1
Overview of data science
70% theory · 30% practice
- Careers: data analyst, data scientist, data engineer
- Lifecycle of a data project
- Types of data and problems (prediction, classification…)
- Ethical issues and data quality
Python for data
30% theory · 70% practice
- Installing the environment and discovering Jupyter
- Variables, types, lists and dictionaries
- Conditions, loops and functions
- Reading and writing files
- Modules and libraries
🛠 Hands-on workshops
First Python programs2 h
A series of progressive exercises to build programming reflexes.
✅ Deliverable : A notebook of solved exercises
Day 2
Object-oriented programming in Python
30% theory · 70% practice
- Why object-oriented programming: structuring and reusing your code
- Classes, objects, attributes and methods; the __init__ constructor and self
- Encapsulation: public and private attributes, properties
- Inheritance, polymorphism and overriding special methods (__str__, __repr__, __eq__)
- Class methods and static methods; data classes (dataclasses)
- Error handling and custom exceptions
- Organizing your code into modules and packages
- Link with data science: classes to load, clean and transform data; the logic behind scikit-learn objects (fit, predict, transform)
🛠 Hands-on workshops
Build a small object-oriented data pipeline3 h
Write classes that load a dataset, clean it and compute indicators, then reuse them in a notebook.
✅ Deliverable : A reusable Python module with its tested classes
Day 3
Handling data with pandas and NumPy
25% theory · 75% practice
- NumPy arrays and vectorized computation
- pandas DataFrames: reading, selecting, filtering
- Cleaning: missing values, duplicates, types
- Grouping, sorting and merging
- Creating new variables
🛠 Hands-on workshops
Clean a real dataset2.5 h
Prepare a raw dataset for analysis: fixing errors, harmonizing, new variables.
✅ Deliverable : A clean, documented dataset
Visualizing data
25% theory · 75% practice
- Principles of good visualization
- Charts with Matplotlib and Seaborn
- Distributions, relationships and comparisons
- Telling a story with your charts
Day 4
Statistics for data science
40% theory · 60% practice
- Descriptive statistics: mean, median, dispersion
- Correlation and causation
- Probability and common distributions
- Notions of sampling and testing
Machine learning: concepts and regression
30% theory · 70% practice
- Supervised and unsupervised learning
- Training and test sets
- Linear regression with scikit-learn
- Preparing variables for a model
🛠 Hands-on workshops
Predict a value2 h
Build a regression model and measure its quality.
✅ Deliverable : An evaluated regression model
Day 5
Classification and model evaluation
30% theory · 70% practice
- Classification: logistic regression, decision trees
- Confusion matrix, precision and recall
- Overfitting and cross-validation
- Comparing several models
🛠 Hands-on workshops
Classify observations1.5 h
Train and compare two classification models on a business case.
✅ Deliverable : A model comparison
Ethics, bias and limits
70% theory · 30% practice
- Bias in data and models
- Interpretability and accountability
- Privacy and compliance
End-to-end data mini-project
10% scoping · 90% implementation
- Choosing a dataset and a question
- Cleaning, exploration, modeling
- Presenting the results
🛠 Hands-on workshops
Mini-project3 h
Each participant completes and presents a full project, from raw data to conclusion.
✅ Deliverable : A project notebook and its presentation
Tools and technologies
Jupyter NotebookMatplotlibNumPyPythonSeabornpandasscikit-learn
Teaching methods
- Short theory sessions followed by immediate hands-on work
- Guided notebooks, then independent exercises
- Real datasets from business cases
- Capstone mini-project and individual reviews
Assessment
- Graded exercises in each module
- Mini-project presented to the group
- Knowledge-check quiz
Required equipment
- A computer with an internet connection
- Python and Jupyter (guided installation at the start of the course) or access to an online environment
Resources provided
- Course materials and solved notebooks
- Exercise datasets
- pandas and scikit-learn cheat sheets
Career outcomes
- Solid foundations to continue towards data analyst or data scientist
- Independence in analyzing data with Python
- Preparation for advanced machine learning courses
Related video courses
My video courses that go with this program, to get ready for the training or to review afterwards. Free with an account.
