Teacher since May 2025
Leandro Ferraz
Calculus and optimisation for machine learning, done by hand
9
tutors built
4.6Sample
average from 18 reviews
319Sample
lessons taught by their tutors
About Leandro
I teach the calculus and optimisation that make models learn: derivatives, gradients, the chain rule, gradient descent and the loss functions it minimises. My background is in engineering and numerical computing, so I care about why things work and also about when they break, such as unstable learning rates or overflowing exponentials. I teach with pencil calculations first, then a few lines of plain pseudocode, never a black box. I am patient with rusty algebra and I will always show the intermediate steps, because skipping them is where most confusion starts.
Knows about
Tutors by Leandro
9 tutors
Logs and Exponentials for ML
Get comfortable with exponents and logarithms, the quiet workhorses of machine learning formulas69 lessonsSampleLeandro FerrazFreeDerivatives and Gradients from ScratchDerivatives and Gradients from Scratch
Understand rates of change, derivatives and gradients as the compass that guides learning62 lessonsSampleLeandro Ferraz$4The Chain Rule and BackpropagationThe Chain Rule and Backpropagation
Compute gradients through a network by hand and see exactly what backpropagation does52 lessonsSampleLeandro Ferraz$7Optimisation Basics for MLOptimisation Basics for ML
Understand momentum, adaptive methods, schedules and the shape of loss landscapes50 lessonsSampleLeandro Ferraz$11Loss Functions and What They RewardLoss Functions and What They Reward
Choose a loss that matches what you actually care about, and know what each one optimises50 lessonsSampleLeandro Ferraz$7Which Maths Do You Need for ML?Which Maths Do You Need for ML?
Build a realistic maths study plan matched to the kind of AI work you want to do36 lessonsSampleLeandro Ferraz$5Gradient Descent by HandGradient Descent by Hand
Take gradient descent steps with a pencil and feel how learning rates make or break trainingLeandro Ferraz$4Numerical Stability: Softmax and Log-Sum-ExpNumerical Stability: Softmax and Log-Sum-Exp
Stop overflows, underflows and NaNs by computing ML formulas the stable wayLeandro Ferraz$10Least Squares from First PrinciplesLeast Squares from First Principles
Derive the normal equations, see regression as projection and solve it the stable wayLeandro Ferraz$10Recent reviews
What students said about Leandro's tutors.
- Ines B.Sample
Checking every rule with a calculator made calculus feel trustworthy rather than magic. Gradients finally make sense.
- Fiona C.Sample
Realistic about my limited hours. The diagnostic questions were quick and pinpointed my gaps.
- Mariana C.Sample
Training curve diagnostics were excellent. Dense material, so I took it slowly over a few weeks.
- Hanna L.Sample
I failed calculus at school. This tutor showed every step and I never felt lost.
- Ayumi S.Sample
The vanishing gradient explanation through the 0.25 sigmoid bound was simple and memorable.
- Daniela K.Sample
Good explanation of why log loss rewards calibration. Would like more on custom losses.