Indie Degree
GitHubSign in
← Programme

Machine Learning Foundations

AIE-202

3 credits · 60h required · after AIE-201

The classical grounding that makes deep learning legible rather than magical, and the source of almost all evaluation intuition. Everything you learned to measure in AIE-102 was invented here, on models small enough to see through. A model you have implemented from the update rule is a model whose failures you can diagnose; one you have only called is one you can only restart.

The lab. The same problems, taken seriously this time. Career Side Quests' requirement-versus-boilerplate line classification is a genuine supervised problem with labels you can produce in an afternoon. Indie Degree's own resource corpus is an unsupervised problem where you already computed an SVD in AIE-201 and can now ask whether the clusters agree with the components. Carpark SG's rate data is a regression problem with real messiness and a ground truth that exists.

By the end you can

  • Implement linear and logistic regression, and a tree ensemble, from scratch
  • Diagnose bias and variance from learning curves
  • Choose and justify a metric for an imbalanced problem
  • Explain what a model is fitting, and to what

What this builds

  • Machine learning foundations · supporting0/17 demonstrated

Next assessment. Linear and logistic regression from scratch — 180 min, tier 1. Machine-checkable, so you can attempt it before reading anything and find out whether you need to.

0/17 assessed items demonstrated · 0/11 read or watched. The second number is how you close a gap; only the first is evidence that it closed.

0 of 17 demonstrated · 0/11 read

0m of 60h

M1 · Supervised learning, from the update rule

0/6 · 14h

M2 · Bias, variance, and what the curves say

0/5 · 12h

  • lecture · 3h · tier 0 self-marked

    Stanford CS229: Machine Learning Course | Summer 2019 (Anand Avati) — Stanford Online

  • reading · 2h 45m · tier 0 self-marked

    Machine Learning Crash Course

  • assignment · 3h · tier 1 machine-verified

  • assignment · 3h · tier 2 panel-assessed

  • retention · 15m · tier 1 machine-verified

M3 · Features, and the ways they lie

0/5 · 10h

M4 · Unsupervised, and its temptations

0/5 · 10h

M5 · Metrics, imbalance, and being believed

0/7 · 14h

  • reading · 2h · tier 0 self-marked

    Metrics and scoring: quantifying the quality of predictions

  • reading · 2h · tier 0 self-marked

    Probability calibration

  • assignment · 3h · tier 1 machine-verified

  • assignment · 2h 45m · tier 2 panel-assessed

  • project · 3h · tier 3 artifact

  • defense · 1h · tier 4 defended

  • retention · 15m · tier 1 machine-verified