Lesson 0003

Action and its fruits

The model is finished. Now: how do you know if it's good — and good for what? This is where knowing stops being enough and someone has to act.

The measure fits the thing measured

A scale weighs; a clock times; neither can do the other's job, and asking a clock "how heavy" is a category error, not a hard question. STOR 323 makes the same point about evaluating models: the metric follows the object being evaluated. A number gets measured by how far off it is (MSE). A ranking gets measured by how well-ordered it is (ROC/AUC). A yes-or-no decision gets measured by which specific errors it makes (the confusion matrix). Using the wrong one doesn't just give a worse answer — it answers a question nobody asked.

MSE you've effectively already met: back in lesson 1, the "error" number that dropped as you tuned the impression-line was exactly this — the average squared distance between the impression and what actually happened. It answers "how far off," and nothing more. The richer question is what comes next.

The threshold is an action, not a fact

A screening test returns a score between 0 and 1 — call it the model's honest best guess. But a number alone refers no one for treatment. Someone has to draw a line: score above this, act; below, don't. That line is a threshold, t, and choosing it is not a discovery about the world. It's a decision.

Primary source — Bhagavad Gītā 2.47
"You have a right to your actions, but never to the fruits of your actions." Karmaṇy-evādhikāras te mā phaleṣu kadācana — Kṛṣṇa's counsel to Arjuna, facing a decision with real consequences he cannot fully control. Bhagavad-gītā 2.47 · comparative translations

Choosing t is precisely this kind of act. You have the right to set the threshold — that part is yours. You do not have the right to which specific patient turns out to be a false alarm or a missed case — the fruits. Move the threshold below to see the shape of the choice:

Notice: you can shift where the errors fall, but you cannot make both kinds of error disappear at once. That tension isn't a flaw in your choice of t — it's a fact about the model's ranking, prior to any action at all. What Gītā 2.47 asks of you is exactly this: commit fully to the choice of t, without the illusion that committing to it lets you also control which case becomes the false alarm.

Stepping back from any one threshold

Every choice of t gives one confusion matrix, one pair of error rates. But there's a question underneath all of them: independent of where you draw the line, is this model any good at ranking the sick above the healthy at all? Sweeping t across its whole range traces the ROC curve; AUC compresses that whole sweep into a single number.

The concept — sākṣī bhāva
In Advaita Vedānta, the sākṣī is the witness — pure awareness that observes the mind's changing states without being any one of them, without being pulled into or identified with a single passing state. Sakshi (witness) — overview

Look at the small plot beside the sliders above: as you swept t, it traced a trail — not one point, the whole curve. That trail is what a witness's view of the model looks like: not attached to any single threshold's verdict, but taking in the entire range of what the model could say. AUC is what remains when you refuse to over-identify with any one decision. It answers "how good is the seeing," not "what was decided" — and, tellingly, it cannot tell you where to set t. That question is still yours; sākṣī bhāva was never a substitute for acting, only for clarity before acting.

One rule, unevenly felt

Here is the sharpest turn in the lecture. Apply the exact same threshold — the exact same rule — to two different groups of patients. Group A: 40% of healthy patients get a false alarm. Group B: 20%. Same model, same t, same intention. Different real burden.

The concept — adhikāra-bheda
Classical Vedānta holds that the same teaching cannot be given identically to every student — adhikāra, readiness or fitness, differs person to person, and what serves one student's understanding may miss another's entirely. A single uniform instruction, applied without regard to differing starting conditions, does not land the same way twice. Adhikari-bheda — overview

Adhikāra-bheda is a claim about teaching, not about statistics — the mechanism here is different (base rates and score distributions, not spiritual readiness). But the structural insight transfers cleanly: uniform treatment, applied without attention to differing starting conditions, does not produce uniform outcomes. "Same rule for everyone" sounds like fairness. It is not the same claim as "same burden for everyone," and STOR 323 states this as plainly as the tradition does: equal rule ≠ equal burden.

Check what's actually landed

Answer from memory before checking.

What's the STOR 323 term for the cutoff — your "right to action" — that turns a score into a decision?
What's the name for the TP/FP/FN/TN grid produced by one fixed threshold — the record of "the fruits" you don't get to choose?
Fill in the blank: equal rule ≠ equal ____

Go deeper

Formal side: An Introduction to Statistical Learning Ch. 4 on ROC curves, and Fairness and Machine Learning for the group-fairness material properly developed. Philosophical side: the Gītā's second chapter in full is worth reading past verse 2.47 — the whole chapter is Kṛṣṇa building the case for right action under uncertainty. See RESOURCES.md.

If adhikāra-bheda felt like the weakest parallel of the three lessons so far, say so directly — I flagged above that the mechanism doesn't match exactly, only the structural insight, and that's a judgment call worth checking with you rather than assuming. Next: a single exercise that puts all three lessons together into one decision, played out step by step.