The Maximum-Entropy Principle and Demand Reconstruction from Missing Data: the Least-Biased Order-Value Distribution Given Only a Mean and Bounds: chapter answers
Owners who open a new shop or launch a new product often know only three numbers: a floor, a ceiling and an average for order value or demand. The usual moves are "treat it as normal" or "if we know nothing, treat it as uniform", both regarded as the least prejudiced choice. This book overturns that common sense. The normal shortcut adds a symmetric shape and a standard deviation; under illustrative parameters, with a mean of 80 and a floor of 30, it puts 0.133 of its probability on impossible values, rising to 0.412 when the mean is 40. The uniform shortcut contradicts the known mean. Accept…
DOI pending · monograph page goes live after publication · 5DL-B-301 · 中文版
- Chapter 1 · The Lagrange Multiplier of Maximum Entropy: the Truncated Exponential When Only a Mean and Bounds Are Known
The closer the mean sits to the lower bound, the worse the normal shortcut becomes. With a mean of 40 it puts 0.412 of its probability below 30, and its 10th percentile is a negative order value of $-17.7$, impossible i… - Chapter 2 · Discrete Maximum Entropy over Price Tiers: Wallis's Count, the Concentration Theorem and the Limits of Uniqueness
Raising the average from 79 to 85 lifts the share at 149 and above from 0.1770 to 0.2100; lowering it to 73 cuts the share to 0.1438. A difference of six in the average shifts the large-order share by three to four perc… - Chapter 3 · The Cost of the Normal and Uniform Shortcuts: Quantiles, Log Loss and the Minimax
Chapter 3 compares the three approaches under three rulers: across four illustrative truths the log loss of maximum entropy is constant at 7.080 bits and smallest in the worst case, yet the worst error of the 90th perce… - Chapter 4 · Adding the Second Moment: the Truncated Normal Family, Convex Exponents and What One Standard Deviation Buys
At standard deviations of 60 and 75 the maximum-entropy answer becomes convex. The 90th percentile is pushed to 166.1 and 236.8 while the median drops to 57.0 and 47.9: polarisation empties the middle. The uncut normal… - Chapter 5 · After the Sample Arrives: Plug-in Bias, the Second-Moment Trap and the Handover Point
Look first at the root mean square error. With only 5 orders, maximum entropy errs by 45.2, the sample quantile by 52.4 and the normal by 50.0. With 10 orders the figures are 34.1, 44.4 and 38.0; with 20 they are 24.8,… - Chapter 6 · Stocking a New Product with the Newsvendor Model: Ordering at a Maximum-Entropy Quantile and Worst-Case Regret
Chapter 6 puts the maximum-entropy quantile into the newsvendor model for new products; for thin-margin lines the normal shortcut's worst regret share reaches 1.018. - Chapter 7 · East and West in the Classics: "Knowing That One Does Not Know" in Laozi 71 and Laplace's Philosophical Essay on Probabilities
Chapter 7 reads "knowing that one does not know" in Laozi 71 (Wang Bi and silk texts) and Laplace's principle of insufficient reason, measuring by divergence the pretended knowledge of distributions that contradict the… - Chapter 8 · Closing the Loop: Cold-Start Demand for a New Product, the Maximum-Entropy Demand Reconstructor and a Pre-Registered Plan of Refutation
Chapter 8 gives the cold-start loop for new products, the distilled Maximum-Entropy Demand Reconstructor and a pre-registered plan of refutation, in which detecting an exceedance share of 0.20 in the calibration test ne…
Illustrative model calculation, not a live quote