Better Life Choices Lab · Decision Framework

-
Better Life Choices Lab · Decision Framework (EN/繁中)
Better Life Choices Lab
A data-driven decision loop: constraints → experiments → reflection

01 · Foundations: a reliable decision loop

This app turns “life choices” into a repeatable practice: define constraints, compare options, test assumptions, and reflect.

Loop
Constraints → Options → Scoring → Experiments → Reflection
Rule of thumb
Prefer reversible moves when uncertainty is high
How to use this page (touch-first)
  • Pick a case, or write your own options.
  • Adjust weights to match your values (not your mood).
  • Look at sensitivity: if rankings flip easily, you need better info.
  • Use the 3D view to spot trade-offs visually.
  • Train bias detection in the mini practice.

02 · Case library (3 diverse starting points)

Cases are not answers. They are templates to reveal constraints and missing information.

Case A · Career switch

High uncertainty, long-term impact. Typical tension: growth vs stability.

A

Case B · Move for relationship

A systems decision: identity + support network + career continuity.

B

Case C · Build a health habit

Low barrier, high compounding. Focus on environment constraints and consistency.

C

Your case · Your options

Start with 2–4 options. If you only have 1, you’re not deciding yet—you’re describing.

Custom

03 · Decision Lab (the practical tool)

Convert ambiguity into a structured comparison. Then attack your structure with sensitivity analysis.

Model

Ratings: 0–10 (higher is better). Weights: 0–10. Risk: 0–10 (higher means more uncertainty penalty).
Options
Criteria
Uncertainty
Export / Import
Copy JSON for backup or sharing. (No downloads, no popups.)

Results

Current best option
—
Decision stability (sensitivity)
—

Ranking

Option Final score Uncertainty penalty
If scores are close, treat it as a signal: you need better information, not better math.

Why this ranking?

Next steps (turn unknowns into tests)

04 · 3D Decision Space (Three.js)

Each option becomes a point in space: Impact (X), Effort (Y), Risk (Z). Tap a point to inspect.

Controls

Touch drag to rotate, pinch to zoom. This view updates from Decision Lab.

Inspector

Tap a point to view details.

05 · Bias Practice (Phaser)

A short practice: classify thought-cards. Goal: notice bias patterns before they drive decisions.

How it works

  • Tap left zone: “Emotion-driven”. Tap right zone: “Logic-driven”.
  • Some cards are mixed—practice naming the bias, not judging yourself.
  • After 60 seconds, you’ll see a pattern summary.

Session stats

Time: —
Correct: —
Misses: —
Most frequent bias: —

06 · Reflection Planner (daily loop)

Better choices are trained through review. Keep it lightweight and consistent.

Daily reflection prompt

—

Reflection log

Tip: look for patterns (e.g., consistently underestimating fatigue or overvaluing short-term relief).

Follow-up questions

  1. How might incorporating daily reflection practices change the way you evaluate potential decisions in uncertain situations?
  2. What alternative strategies could you use if emotional biases consistently influence your choices despite logical analysis?
  3. In what ways could reevaluating past decisions through a structured framework reveal patterns that improve future outcomes?
Ready.
Single-file demo · Touch-first · No external assets · Three r134 + Phaser 3.90.0

此網誌的熱門文章

自訂網路結構的神經網路訓練與預測 (動畫＋公式＋損失／激活函數

-
自訂網路結構的神經網路訓練與預測 (動畫＋公式＋損失／激活函數) 自訂網路結構的神經網路訓練與預測 此範例示範一個多層神經網路，其網路結構、激活函數及損失函數可依需求自訂。 訓練與預測均以動畫展示前向傳播（並於訓練時進行反向傳播更新權重），搭配公式與統計資訊。 深入了解神經網路的基本元素與計算方法 在本範例中，我們介紹神經網路中的各個核心基本元素，如參數與超參數，並說明它們如何在模型計算與實際應用中發揮作用。 參數 (Parameters) 參數主要指神經網路中的權重 (Weights) 與偏置 (Biases)。在前向傳播時， 輸入資料依據公式： a = activation(Σ(x * weight) + bias) 其中，x 為輸入值、weight 為權重、bias 為偏置，activation 則代表激活函數，用以引入非線性。權重與偏置在訓練過程中透過反向傳播 (Backpropagation) 依據損失函數的梯度進行更新，以達到模型調整誤差的效果。 超參數 (Hyperparameters) 超參數是在訓練前由使用者設定的參數，不會在訓練過程中自動調整。常見超參數包括： 學習率 (Learning Rate)： 控制每次參數更新的步幅，影響收斂速度與穩定性。 Epoch 次數： 表示整個資料集被完整遍歷的次數，決定訓練強度與時間。 網路結構設定： 包含輸入層、隱藏層與輸出層神經元的數量，直接決定模型表達能力。 激活函數與損失函數： 分別決定每層的非線性處理與模型誤差的計算方式。 合理配置超參數能夠有效避免過擬合或欠擬合，並確保網路運算的穩定與高效。舉例來說，若學習率過高，可能會使參數更新過度，導致模型難以收斂；相反，學習率過低時則會使訓練變慢。 基本原理與實際應用 神經網路的運算流程主要分為前向傳播與反向傳播兩大步驟： ...
閱讀更多

Customizable Neural Network Training and Prediction (Animations + Formulas + Loss/Activation Functions)

-
Customizable Neural Network Training and Prediction (Animations + Formulas + Loss/Activation Functions) Customizable Neural Network Training and Prediction This example demonstrates a multi-layer neural network where you can customize the network structure, activation functions, and loss functions as needed. Both training and prediction are shown with animations illustrating forward propagation (with back propagation update during training), accompanied by formulas and statistical information. In-depth Understanding of Neural Network Basic Elements and Calculation Methods In this example, we introduce the core components of a neural network—parameters and hyperparameters—and explain how they function in model computation and practical applications. Parameters Parameters refer mainly to the weights and biases within the network. During forward propagation, input data is computed as: a = activa...
閱讀更多