Instrumentation & Measurement Magazine 24-2 - 91

Table 2 - Comparing I&M and ML terminologies (continued)
ML regression
Term

Uncertainty

meaning in I&M

parameter characterizing the
dispersion of measurements
[11, 2.26]; can be evaluated as
Type A or Type B [14]

meaning

ML classification
relation with I&M

both recognize the fact
that repeated actions; i.e.,
measurements of the same
measurand in I&M and
estimations of the same input
in variable-output ML, give
a different result each time.
Note that I&M uses standard
deviation; i.e.,  Variance , while
ML uses Variance, to quantify
uncertainty

measure of the reliability of a
variable-output ML model's
estimation [18][19]; does not
apply to fixed-output ML
models

relation with
I&M

meaning

same as ML regression

✔

✔

Epistemic
uncertainty

not an official term

the limited/incorrect
knowledge can be caused by
I&M systematic error during
the measurement of the
training data, but also by the
unavailability of data in certain
input ranges. See Fig. 2 in Part
2 [2] for more details.

model uncertainty due to the
limited/incorrect knowledge
that the training data provides
to the ML model; can be
reduced by collecting more/
better data [20]

same as ML regression

✔

Aleatoric
uncertainty

not an official term

uncertainty caused by the
fact that the individual data
points collected for the training
dataset have random noise;
cannot be reduced by collecting
more data [20]

✔

the training dataset's
random noise is the result
of I&M random error in the
measurement of those data

✔

Table legend
TP: number of true positives. A true positive is the correct classification of an actual positive as positive.
TN: number of true negatives. A true negative is the correct classification of an actual negative as negative.
FP: number of false positives. A false positive is the incorrect classification of an actual negative as positive.
P: number of actual positives
N: number of actual negatives
✔ conceptual similarity but not identical
✔ no similarity but some relation
✘ no similarity and no relation

process without an experimental stage is not officially measurement [22, p. 203], this class of applications is not officially
considered to be measurement. Nevertheless, it's important
to mention this class of applications, because with the advent
of ML, we expect to see more and more applications like this
which predict a measurement result without actual experimental measurement.
In Part 2 [2], we will show how the uncertainty introduced
by the ML module of a measurement system can be quantified.
The customary Conclusion section will also be given in Part 2.

(IEV) - Part 300: Electrical and Electronic Measurements and
Measuring Instruments, International Electrotechnical
Commission, Jul. 2001.
[5]	 P. Pouladzadeh et al., " You are what you eat, so measure what you
eat! " IEEE Instrum. Meas. Mag., vol. 19, no. 1, pp. 9-15, Feb. 2016.
[6]	 M. Vallejo, C. de la Espriella, J. Gómez-Santamaría, A. F. RamírezBarrera, and E. Delgado-Trejos, " Soft metrology based on
machine learning: a review, " Meas. Sci Technol., vol. 31, no. 3, Mar.
2020.
[7]	 H. Goldberg, " What is virtual instrumentation? " IEEE Instrum.

References

Meas. Mag., vol. 3, no. 4, pp. 10-13, Dec. 2000.

[1]	 M. Khanafer and S. Shirmohammadi, " Applied AI in
instrumentation and measurement: the deep learning
revolution, " IEEE Instrum. Meas. Mag., vol. 23, no. 6, Sep. 2020.
[2]	 H. Al Osman and S. Shirmohammadi, " Machine learning in
measurement, part 2: uncertainty quantification, " IEEE Instrum.
Meas. Mag., vol. 24, no. 3, 2021 (in press).

[8]	 S. Shirmohammadi and A. Ferrero, " Camera as the instrument:
the rising trend of vision based measurement, " IEEE Instrum.
Meas. Mag., vol. 17, no. 3, pp. 41-47, Jun. 2014.
[9]	 P. Pouladzadeh et al., " Measuring calorie and nutrition from food
image, " IEEE Trans. Instrum. Meas., vol. 63, issue 8, pp. 1947-1956,
Aug. 2014.

[3]	 D. Petri, " Big data, dataism and measurement, " IEEE Instrum.
Meas. Mag., vol. 23, no. 3, pp. 32-34, May 2020.
April 2021	

[4]	 IEC standard 60050-300, International Electrotechnical Vocabulary

[10]	F. P. W. Lo, Y. Sun, J. Qiu and B. Lo, " Image-based food
classification and volume estimation for dietary assessment: a

IEEE Instrumentation & Measurement Magazine	91
Instrumentation & Measurement Magazine 24-2

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-2
