Backporch
Tutoring & collaboration

Learn it properly — mathematics & data science, one-on-one.

Teaching is the work I want to do for a lifetime. Patient, concept-first, and paced to you — from a single tough homework set to the whole arc of becoming a data scientist. Any student is welcome; if you're heading into data science, I'd especially love to help.

Schedule a sessionOnline, flexible, first conversation is free.

Who I work with

Students, any level

High school through graduate — homework help, exam prep, or genuine understanding that outlasts the test.

Your on-ramp: Foundations phase I

What we can cover

Mathematics
From algebra and calculus to linear algebra, probability, and proof — including the math that quietly sits under every model.
Statistics
Intro through mathematical statistics: regression, experimental design, and Bayesian thinking — plus the traps that make smart people wrong.
Operations research & optimization
The algorithms for deciding under constraints: linear and integer programming, the simplex and interior-point methods, network flows, shortest-path and assignment, scheduling and routing, queueing theory, dynamic programming, and metaheuristics (simulated annealing, genetic algorithms). Formulating a real problem and choosing the method that fits — not just calling a solver.
Decision Mechanics
The wider toolkit under a decision — spectral and matrix methods, Monte-Carlo simulation, Markov models, and game theory — built and understood from the ground up.
Analytics & data science
Turning data into decisions: Python, SQL and data modeling, forecasting, dashboards (Power BI, Tableau), and the whole workflow from question to recommendation.

The data-science track

For anyone aiming at a data-science degree or career, this is a deliberate build — the whole stack, in order, so nothing stays a black box. But it's one spine with three on-ramps: you start where you already are, not always at step one. Every stop links to a live demo on this site, so you can see how we'd actually learn it before we ever meet.

  1. New to it all? Start here.
    I
    Foundations

    the black boxes, opened

  2. Already code? Join here.
    II
    Core

    how models actually work

    • Probability & distributions
      Where distributions come from and why the Gaussian shows up everywhere — built, not memorized.
      distribution foundry
    • Statistics & inference
      From regression and hypothesis testing to the honest evaluation of a result — what it means, and when it doesn't.
      law of large numbers
  3. Using the tools already? Start here.
    III
    Applied

    turning data into decisions

    • Modeling & regularization
      Fitting a model without fooling yourself — bias, variance, and generalization you can defend.
      scaling & regularization
    • Optimization & decisions
      Choosing under hard constraints when there's no exact answer — the honest heuristics, and their limits.
      the traveling salesman
  4. IV
    Portfolio

    something you can defend

    • From coursework to a project
      Turn assignments and Kaggle-style problems into work you can defend in an interview — the whole ML workflow, the reasoning, not just the score.
      the ML pipeline, end to end

How I teach — the method behind the sessions

One belief runs through all of it — formulation over solution, from how I think: get the shape of a problem right and the answer is half-forced. That's also how I learn, and how I'll teach you. Skill is built, not waited for: higher-order encoding and mind-mapping (in the spirit of Justin Sung), then retrieval, spacing, and teaching it back. You won't just get answers — you'll leave with a way to build durable understanding on your own.

Encode, don't collect
Notes you only gather slide off; notes organized by relationship stick.
how Before writing anything down, draw the shape — a mind-map of what connects to what. The structure is the memory.
Retrieve, then space it
Pulling an idea from memory strengthens it far more than reading it again.
how Close the book and reconstruct from nothing; then revisit on widening intervals — let a deck keep the schedule.
Reconstruct before you check
The struggle to rebuild surfaces the gaps that re-reading quietly hides.
how Re-derive the result cold, from the idea alone — only then compare against the page.
Teach it to find the gaps
Explaining it simply exposes what you only thought you understood.
how Write it for a beginner. Every stumble is a gap with an address — go fix that one.

A worked example — watch one topic encode

The first move on a single idea — L² geometry — as a live mind-map that builds itself, hub first and then one relationship at a time, because that ordering is the encoding. Tap a branch; several open the proof that makes the relationship real. This is the move we'd do together on whatever you're studying.

L²geometryAngleLengthProjectionDistanceSpectralρ = cos θCauchy–Schwarz‖X‖² =Var(X)standard-izeleastsquarescond.expectationEuclid &Mahalanobisp-normseigen →SVD → PCAwhitening
One idea in the middle; the branches are relationships, not a list; the dashed cross-links are where those relationships turn out to be the same idea. Tap a branch for its role, or a leaf to open the proof that makes the link real.

Research collaboration

Beyond tutoring, I collaborate on quantitative research across the sciences — the statistics and modeling side of an honest question. If you work in one of these (or somewhere adjacent), I'd like to compare notes.

Epidemiology & biostatistics
Epidemic and clinical data — incidence, survival, hierarchical and spatial models.
Policy & polling
Survey design, weighting, and social-behavioral measurement — reading a population honestly.
Psychology & cognitive science
Measurement of mind: psychometrics, experimental design, and the statistics of contested claims.
Neuroscience & neural computation
Spiking and population models, neural data, and the computation behind cognition.
Astrostatistics
Inference on the sky — faint signals, selection effects, and uncertainty at cosmic scale.
Statistical mechanics
Ensembles, entropy, and the bridge from micro-rules to macro-behavior — physics as inference.
How it works. We start with a free conversation about where you are and what you want. Sessions are online and flexible; I'll suggest a cadence and, for students, work with your coursework rather than around it. Reach me at tutoring@backporch.studio for tutoring, or research@backporch.studio for research collaboration.