explain xkcd - User contributions [en]

Talk:2048: Curve-Fitting

2018-09-19T18:00:44Z

Werhdnt:

> House of Cards: Not a real method, but a common consequence of mis-application of statistical methods: a curve can be generated that fits the data extremely well, but immediately becomes absurd as soon as one glances outside the training data sample range, and your analysis comes crashing down "like a house of cards". This is a type of _overfitting_

I'm pretty sure it refers to the TV show house of cards, the dots representing the quality of the series increasing until Netflix renewed it a bit too much

I'm a little mystified by the alt-text. Cauchy and Lorentz both seem like mathematically capable people. What am I missing? [[Special:Contributions/172.69.62.226|172.69.62.226]] 17:46, 19 September 2018 (UTC)

: Google-Fu reveals that it's a continuous probability distribution. This isn't bad per se, but it is quite visually distinctive and also can be quite...concerning if the data set isn't one where probability should be an issue. [[User:Werhdnt|Werhdnt]] ([[User talk:Werhdnt|talk]]) 18:00, 19 September 2018 (UTC)

Talk:2048: Curve-Fitting

2018-09-19T18:00:25Z

Werhdnt:

2048: Curve-Fitting

2018-09-19T17:56:41Z

Werhdnt: /* Explanation */

{{comic
| number = 2048
| date = September 19, 2018
| title = Curve-Fitting
| image = curve_fitting.png
| titletext = Cauchy-Lorentz: "Something alarmingly mathematical is happening, and you should probably pause to Google my name and check what field I originally worked in."
}}

==Explanation==
{{incomplete|Please edit the explanation below and only mention here why it isn't complete. Do NOT delete this tag too soon.}}

A illustration of several plots of the same data with curves fitted to the points, paired with conclusions that you might draw about the person who made them.

When modeling a phenomenon statistically, it is common to search for trends, and fitted curves can help reveal these trends. Much of the work of a data scientist or statistician is knowing which fitting method to use for the data in question.

In general, the researcher will specify the form of an equation for the line to be drawn, and an algorithm will produce the actual line.

* Linear: f(x) = mx + b

* Quadratic: f(x) = ax^2 + bx + c

* Logarithmic: f(x) = a*log_b(x) + c

* Exponential: f(x) = a*b^x + c

* Loess: w(x) = (1-|d|^3)^3

* Linear, No Slope: f(x) = c

* Logistic: f(x) = L / (1 + e^(-k(x-b)))

* Confidence Interval: not a type of curve fitting, but a method of depicting the predictive power of a curve

* Piecewise: Mapping different curves to different segments of the data. This is a legitimate strategy, but the different segments should be meaningful, such as if they were pulled from different populations.

* Connecting lines: Not useful whatsoever, but it looks nice!

* Ad-Hoc Filter: Drawing a bunch of different lines by hand. Also not useful.

* House of Cards: Not a real method, but a common consequence of mis-application of statistical methods: a curve can be generated that fits the data extremely well, but immediately becomes absurd as soon as one glances outside the training data sample range, and your analysis comes crashing down "like a house of cards". This is a type of _overfitting_

* Cauchy-Lorentz: Is a continuous probability distribution, which might be rather concerning if the data set is something like 'the heights of the people in this building.'

==Transcript==
{{incomplete transcript|Do NOT delete this tag too soon.}}
:'''Curve-Fitting Methods'''
:and the messages they send

:[In a single frame twelve scatter plots with unlabeled x- and y-axes are shown. Each plot consists of the same data-set of approximately thirty points located all over the plot but slightly more distributed around the diagonal. Every plot shows in red a different fitting method which is labeled on top in gray.]

:[The first plot shows a line starting at the left bottom above the x-axis rising towards the points to the right.]
:Linear
:"Hey, I did a regression."

:[The second plot shows a curve falling slightly down and then rising up to the right.]
:Quadratic
:"I wanted a curved line, so I made one with Math."

:[At the third plot the curve starts near the left bottom and increases more and more less to the right.]
:Logarithmic
:"Look, it's tapering off!"

:[The fourth plot shows a curve starting near the left bottom and increases more and more steeper to the right.]
:Exponential
:"Look, it's growing uncontrollably!"

:[The fifth plot uses a fitting to match many points. It starts at the left bottom, increases, then decreases, then rapidly increasing again, and finally reaching a plateau.]
:Loess
:"I'm sophisticated, not like those bumbling polynomial people."

:[The sixth plot simply shows a line above but parallel to the x-axis.]
:Linear, no slope
:"I'm making a scatter plot but I don't want to."

:[At plot #7 starts at a plateau above the x-axis, then increases, and finally reaches a higher plateau.]
:Logistic
:"I need to connect these two lines, but my first idea didn't have enough Math."

:[Plot #8 shows two red lines embedding most points and the area between is painted as a red shadow.]
:Confidence interval
:"Listen, science is hard. But I'm a serious person doing my best."

:[Plot #9 shows two not connected lines, one at the lower left half, and one higher at the right. Both have smaller curved lines in light red above and below.]
:Piecewise
:"I have a theory, and this is the only data I could find."

{{comic discussion}}

[[Category:Comics with color]]
[[Category:Scatter plots]]
[[Category:Math]]
[[Category:Science]]

Talk:2030: Voting Software

2018-08-08T20:23:10Z

Werhdnt:

I think this comic is referencing [https://twitter.com/GossiTheDog/status/1026603800365330432 this twitter thread] and the controversy behind it.
[[Special:Contributions/172.69.190.4|172.69.190.4]] 17:59, 8 August 2018 (UTC)
: The [https://arstechnica.com/tech-policy/2018/08/experts-criticize-west-virginias-plan-for-smartphone-voting/ Experts criticize West Virginia’s plan for smartphone voting] article on ArsTechnica has more information (as much as possible when the company in question does not provide any details (note that it is about overseas voting). --[[User:JakubNarebski|JakubNarebski]] ([[User talk:JakubNarebski|talk]]) 19:44, 8 August 2018 (UTC)

Is he saying it's weird that we're so sophisticated in other areas of computer science but so far behind in voting technology, or is he making fun of the idea that electronic voting is somehow inherently unsafe?--[[Special:Contributions/108.162.216.106|108.162.216.106]] 18:10, 8 August 2018 (UTC)
: No he is saying computer science is a mess and we should not trust it with voting(he is not making fun of the idea of it being unsafe, he is pressing on the point of it being unsafe[saying that all experts agree on that])18:18, 8 August 2018 (UTC)

: I think he's commenting on how in most fields, the experts are very sure that they do their job well, and all the angles have been tried and tested, but in computer science the experts are more certain than anyone that there is ''absolutely no way'' for a person to actually build a complex software system with no flaws or vulnerabilities, even if they controlled every aspect of the system. in practice of course they control very little of the system and understand even less of it. [[Special:Contributions/172.68.34.88|172.68.34.88]] 18:22, 8 August 2018 (UTC)

: He's saying that software development is a terribly buggy process, most likely because the majority of software out there can have bugs without very dire real-world consequences (unlike aircraft or elevators).

::Not to mention the fact that there are incredibly smart people with great interest in undoing the work that software developers do, whereas that isn't at all the case with airplanes or elevators. [[Special:Contributions/108.162.219.214|108.162.219.214]] 18:29, 8 August 2018 (UTC)

:: Plus there's the general issue that the public as a whole takes the view that "Computers are majykal" (misspelling deliberate) and therefore somehow automatically safe & infallible, despite experts trying very hard to disillusion people about...pretty much all of that. Compare that to the common assumptions about aircraft and elevators--people need the safety verified, instead of assuming it like they do with computers. [[User:Werhdnt|Werhdnt]] ([[User talk:Werhdnt|talk]]) 19:08, 8 August 2018 (UTC)

:::There's a logical fallacy here. To compare airplaneS and elevatorS to a voting system program is comparing plural to singular. There would be significant opportunity to break/modify a single instance of those objects, although without the relative anonymity of electronic access involved. Once a computer system is infiltrated, the break-in can be replicated to all instances of that program relatively instantaneously, assuming communication pathways are available.[[Special:Contributions/162.158.75.130|162.158.75.130]] 19:12, 8 August 2018 (UTC)

::::No logical fallacy; there have been ''multiple'' attempts to get people to accept a voting system program, and the 'done by a computer=infallible' problem is '''''not''''' unique to voting programs. Mr. Babbage was being confused by people who were thinking it was possible to get the correct answers from a computer despite putting the wrong data in back in the 1860s (at least!), and the computer at the time was not much more than a fancy calculator. [[User:Werhdnt|Werhdnt]] ([[User talk:Werhdnt|talk]]) 20:23, 8 August 2018 (UTC)

A blockchain node doesn't technically need to be connected to the internet in order to function. It needs to have some method for receiving messages from other nodes on the blockchain network, and most blockchain nodes do indeed get these messages via the internet, but some bitcoin nodes (for example) get updates about new blocks and new transactions from the Blockstream satellite. An internet connection is therefore not intrinsically necessary for a blockchain to work, it's just the most convenient way to do it.

Do you think that this comic had anything to do with the debacle in Johnson County, KS last night? [[Special:Contributions/162.158.62.231|162.158.62.231]] 19:30, 8 August 2018 (UTC)

Talk:2030: Voting Software

2018-08-08T19:08:20Z

Werhdnt: