为何按公式计算的X1标准化结果0.31被判错误？

Hey Johnny, let's walk through this step by step—your Octave code and math look solid at first glance, so let's unpack why the grading system might have flagged your 0.31 result.

First, let's confirm your calculations are accurate using your provided code and data:

X1 = 7921
midterm_avr = mean([7921,5184,8836,4761])  % = 6675.5
midterm_range = range([7921,5184,8836,4761])  % = 8836 - 4761 = 4075
poly_regress = (X1 - midterm_avr) / midterm_range  % (7921 - 6675.5)/4075 ≈ 0.30564
sprintf('%.*f', 2, poly_regress)  % rounds to 0.31

Your output:
X1 = 7921
midterm_avr = 6675.5
midterm_range = 4075
poly_regress = 0.30564
ans = 0.31

Your math here is 100% correct—you computed the mean, range, and standardized value properly, then rounded to two decimal places as instructed. So why the mismatch? Here are the most likely culprits:

• Different normalization formula expected
  Some grading systems use a variant of standardization. For example:

  • Maybe it expects min-max scaling instead of mean-centered scaling: (X1 - min)/(max - min) would be (7921-4761)/4075 ≈ 0.775, which doesn't match, but if it was (max - X1)/(max - min) that'd be ~0.225—still not 0.31.
  • Or it might use standard deviation instead of range for scaling: The sample standard deviation of your dataset is ~2009, so (7921-6675.5)/2009 ≈ 0.62, which also doesn't align.

• Mismatched dataset
  It’s possible the official midterm dataset used for grading isn’t the four values you used. For example, if there are additional data points, the mean and range would change. Let’s reverse-engineer your accepted X2 result (-0.37):
  If (X2 - avr)/range = -0.37, then X2 = avr - 0.37*range. If your X2 was, say, 5184, solving for avr and range gives 5184 = avr - 0.37*range—a different pair than your 6675.5/4075. This suggests the grading system is using a different set of midterm scores.

• Precision rules (truncation vs rounding)
  If the system expects truncating to two decimals instead of rounding, your 0.30564 would become 0.30 instead of 0.31. That’s a common gotcha—some auto-graders are strict about whether they want rounded or truncated values, even if the difference is tiny.

• Sign error in the formula
  Double-check if the problem specified (midterm_avr - X1)/range instead of (X1 - midterm_avr)/range. That would give ~-0.31, but since your X2 result was negative and accepted, this is less likely unless X2’s formula was flipped too.

My bet is either a dataset mismatch or a precision rule difference. If you can access the official midterm dataset used for the problem, cross-checking that would clear things up immediately.

内容的提问来源于stack exchange，提问作者Johnny