Leçons — Collège

Ces leçons sont en anglais, racontées par les tuteurs Koydo.

Toutes 3d Graphics Rendering Adaptive Mesh Refinement Advanced Algorithms Advanced Software Architecture Algebraic Algorithms Algorithm Design Advanced Algorithmic Game Theory Algorithms Anthropology Api Design Applied Cryptography Applied Machine Learning Approximation Algorithms Ar Vr Systems Assembly Language Astronomy Binary Exploitation Bioinformatics Bioinformatics Algorithms Blockchain Security Circuit Complexity Cloud Computing Cloud Systems Communication Complexity Communications Compilers Computational Chemistry Computational Complexity Computational Electromagnetics Computational Fluid Dynamics Computational Genomics Computational Materials Computational Neuroscience Computational Physics Computational Proteomics Computer Architecture Computer Engineering Computer Graphics Computer Networks Computer Vision Criminal Justice Dance Data Analysis Computing Data Mining Data Science Foundations Deep Learning Dependent Types Disability Studies Education Environmental Policy Food Studies Geology Global Studies Graphic Design Human Nutrition Information Systems Journalism Korean Linguistics Materials Engineering Music Performance Peace Conflict Studies Philosophy Of Religion Political Science Portuguese Public Administration Queer Studies Social Entrepreneurship Sociology Of Health Theater

Introduction to Machine Learning and Mathematical Foundations

Applied Machine Learning · avec South Asian F

Bright university study space bathed in warm sunlight, where a cheerful illustrated computer screen displays a clean diagram showing four colorful pillars labeled "Linear Algebra," "Probability," "Optimization," and "Statistics" supporting a glowing arch labeled "Machine Learning." Surrounding the screen, friendly icons show a matrix grid, a bell curve, a minimizing parabola, and a bar chart. Open textbooks and coffee cups sit nearby on a clean white desk.

Introduction to Neural Networks and Deep Learning

Applied Machine Learning · avec Western Euro F

Bright university study space with large windows letting in golden daylight. A clean whiteboard displays a clear neural network diagram: three labeled columns — "Input Layer," "Hidden Layer," "Output Layer" — with cheerful colored circles (nodes) connected by neat lines labeled with small "w" weights. A friendly arrow curves backward labeled "Backpropagation." Scattered textbooks, a laptop showing a loss curve trending downward, and a coffee mug sit on a tidy desk nearby.

Advanced Optimization: Regularization Techniques

Applied Machine Learning · avec East African F

A bright university classroom whiteboard scene showing two clean graphs side by side under warm daylight. Left graph labeled "L1 Lasso" shows a diamond-shaped constraint region with a cost function ellipse touching at a corner axis point, producing a sparse solution. Right graph labeled "L2 Ridge" shows a circular constraint region with smooth tangent contact. Color-coded penalty terms "+λ|w|" and "+λw²" float above each diagram in bold teal and orange markers.

Natural Language Processing Techniques in Machine Learning

Applied Machine Learning · avec Latina F

A bright, airy university lab with large sunlit windows. A cheerful student points at a glowing whiteboard showing a colorful 3D vector space with labeled word dots — "king," "queen," "man," "woman" — connected by neat arrows illustrating semantic relationships. Nearby, a clean desk displays pipeline stages: "Raw Text → Tokens → Vectors" with small illustrated blocks. Word2Vec and GloVe appear as friendly labeled icons. Everything is crisp, blue-and-yellow toned, energetic, and academically optimistic.

Linear Algebra Basics for Machine Learning

Applied Machine Learning · avec Latina F

A bright, sunlit university classroom whiteboard displays a clean linear algebra visual: two colorful arrows (vectors **a** and **b**) pointing from the origin on a crisp grid, with a dot product formula **a·b = 12** labeled nearby. Beside it, a cheerful 3×3 matrix grid with small numbered cells transforms a cluster of colorful data-point dots into a new organized arrangement, illustrated with a bold curved arrow showing the transformation.

Vector Spaces and Subspaces: Foundations for Machine Learning

Applied Machine Learning · avec East African F

Bright university study room with large sunny windows and white walls. A cheerful student at a clean desk sketches a 3D coordinate grid on graph paper, showing a bold blue plane slicing through the space as a subspace. Colorful arrows represent vectors lying within the plane, labeled **v₁**, **v₂**. A small chalkboard beside shows **αv + βw ∈ W**, illustrating closure. Potted plants and warm sunlight fill the background.

Model Evaluation Metrics and Techniques

Applied Machine Learning · avec Latina F

A bright university lab setting with a large whiteboard displaying a clean confusion matrix grid labeled "TP / FP / FN / TN" in bold colors. A student points enthusiastically at a smooth ROC curve plotted on a sunny yellow chart, with AUC shaded in teal. Nearby, sticky notes show formulas: "F1 = 2×(P×R)/(P+R)" and "Accuracy = Correct/Total." Colorful bar charts and precision-recall diagrams decorate the walls in coral, blue, and green.

An Introduction to Probability Theory for Machine Learning

Applied Machine Learning · avec Western Euro F

A bright, sunlit university classroom with a large whiteboard displaying a clean probability distribution curve (bell-shaped normal distribution) labeled with μ and σ. A confident student points to a colorful diagram showing three connected concepts: a coin flip (binary/classification), a smooth regression curve, and a dice (random variable). Scattered textbooks reference "Laplace." Color-coded probability bars in blue, orange, and green fill a histogram beside the curve. Cheerful, organized, daylight atmosphere.

Recurrent Neural Networks and Time Series Analysis

Applied Machine Learning · avec South Asian F

A bright, sunlit classroom workspace features a cheerful diagram of an RNN unrolled across a whiteboard. Three connected blue nodes labeled t-1, t, t+1 sit in sequence, linked by curved arrows showing information flowing forward. A small clock icon above each node represents memory/hidden state. Below, a clean line graph shows a wavy time series curve in orange. A coffee mug and open notebook with the label "hidden state h" sit nearby.

Conditional Probability and Bayes' Theorem in ML

Applied Machine Learning · avec South Asian F

Bright university classroom with large whiteboard displaying the clean formula **P(A|B) = P(B|A)·P(A) / P(B)** in bold blue marker. A cheerful student points to labeled arrows connecting three glowing circles: "Prior," "Likelihood," and "Posterior" on a sunny yellow diagram. Colorful Venn diagrams overlap in teal and coral on a side panel. A laptop shows a simple spam-vs-email bar chart. Natural daylight streams through tall windows.

Linear Transformations and Applications in Machine Learning

Applied Machine Learning · avec South Asian F

A bright university study space with large sunny windows. A cheerful student stands at a whiteboard sketching two coordinate grids side by side connected by a bold arrow labeled **T(x) = Ax**. The left grid shows a blue square shape; the right grid shows the same shape stretched and rotated into a parallelogram in orange. Nearby, a laptop displays a clean dataset table transforming into normalized values. Color-coded matrix brackets frame the action clearly.

Key Probability Distributions in Machine Learning

Applied Machine Learning · avec East African F

Bright university study space with large sunlit windows. A student at a clean white desk examines four labeled distribution curve cards arranged like a hand of playing cards: a sharp single-bar "Bernoulli p=0.7," a stepped "Poisson λ=3," a multi-colored bar "Multinomial," and a classic bell-curve "Gaussian μ,σ." A laptop displays a colorful scatter plot. Books and coffee nearby. Warm daylight fills the cheerful, organized room.

Understanding Systems of Linear Equations in ML Contexts

Applied Machine Learning · avec Latina F

A bright university classroom with large windows letting in warm sunlight. A friendly whiteboard displays a clean 3×3 matrix equation **Ax = b** with colorful row arrows showing Gaussian elimination steps — rows swapping and combining in blue, green, and orange chalk. Three solution types are sketched as small labeled diagrams: one intersection point (unique), parallel lines (none), and overlapping lines (infinite). A small portrait labeled "Gauss" smiles nearby.

Statistical Inference and Hypothesis Testing in ML

Applied Machine Learning · avec Latina F

A bright university statistics lab with large windows letting in afternoon sunlight. A cheerful student at a clean whiteboard draws a clear bell curve labeled "H₀" with a shaded rejection region marked "α = 0.05." Sticky notes show "p-value," "Type I Error," and "Type II Error" in bold colors. A open textbook displays Ronald Fisher's portrait. Sample data points and a small population diagram connect with confident arrows, illustrating inference flow.

Model Deployment and Real-World Considerations

Applied Machine Learning · avec Western Euro F

A bright, sunlit tech workspace showing a clean architectural diagram on a large whiteboard: a cheerful pipeline flowing from a "MODEL" box through an "API" gateway into three colorful microservice bubbles labeled "Scale," "Monitor," and "Deploy." A confident student points to a "Data Drift?" alert flag while a laptop displays a live accuracy graph trending upward. Blue arrows connect components against a crisp white background with warm yellow accents.

Future Trends in Machine Learning and AI

Applied Machine Learning · avec South Asian F

Bright university research lab bathed in warm afternoon sunlight. A large transparent glass panel displays a glowing neural network diagram with labeled boxes: "Input," "Hidden Layers," "Output," with arrows and small magnifying glasses indicating explainability. Diverse students in casual attire point enthusiastically at the diagram. Sticky notes read "XAI," "Transfer Learning." Clean white desks hold open laptops showing colorful model visualizations. Potted plants and large sunny windows complete the optimistic, modern setting.

Matrices and Matrix Operations in Machine Learning

Applied Machine Learning · avec Western Euro F

Bright university study space with large whiteboard showing two colorful matrices labeled **A** and **B** being multiplied, arrows tracing row-by-column dot products highlighted in orange and teal. A student points enthusiastically at the result matrix **C**. Nearby, a clean identity matrix **I** glows softly in yellow, labeled "neutral element." Sunlit room, crisp clean lines, bold primary colors on white background, textbooks and laptops open.

Ethics and Impact of Machine Learning in Society

Applied Machine Learning · avec East African F

A bright, sunlit university seminar room where a diverse group of students gathers around a large circular table. At the center, a cheerful infographic poster displays three interconnected icons labeled "Fairness," "Accountability," and "Transparency" with colorful arrows linking them. On a whiteboard behind, simple illustrated figures represent healthcare, recruitment, and justice scales. Sticky notes in yellow, pink, and blue surround a dataset diagram highlighting diverse faces and balanced data columns.

Feature Engineering: The Key to Successful ML Models

Applied Machine Learning · avec South Asian F

A bright, sunlit workspace where a confident student transforms a cluttered raw data table into polished, glowing feature columns using a large wrench and paintbrush. Labeled columns show "age," "income," and a newly crafted "debt_ratio" feature highlighted in yellow. A PCA diagram nearby compresses many arrows into two clean principal component axes. Sticky notes reading "domain knowledge" and "select relevant!" decorate the whiteboard. Clean, colorful, optimistic atmosphere.

Convolutional Neural Networks for Vision Tasks

Applied Machine Learning · avec East African F

A bright, sunlit computer science classroom where a large poster on the white wall shows a colorful cat photo being processed through labeled CNN layers: "Conv Layer," "Pooling," "Feature Maps." Friendly illustrated filters slide across pixel grids like magnifying tiles, extracting edges, then shapes, then ears and whiskers. Glowing feature maps shrink progressively in a clean left-to-right pipeline. Students point enthusiastically at increasingly abstract pattern tiles leading to a confident "CAT ✓" label.

Optimization Techniques in Machine Learning

Applied Machine Learning · avec Western Euro F

A bright, sunlit university classroom with a large whiteboard showing a smooth 3D bowl-shaped cost function curve in blue. A cheerful student figure places a small red ball rolling downhill along the curve toward the minimum point labeled "optimal parameters." Arrows indicate gradient descent steps in orange. Clean axis labels read "Loss" and "Parameters." Mathematical notation ∂L/∂w appears neatly beside the curve. Warm yellow light fills the room.