Forecasting & Results - Analysis
#1
Looking at the Table below there have been (5) Forecasts made F1 - F5 and Resulting from the (5) Forecasts made there are (5) Results R1 - R5

So each of the Forecasts F1 - F5 Have some History shown as the Results Numbers 1 - 5

The Question / Problem is what Value/s should R6 be for each of them ? using only the Numbers 1 - 5 ?  Smile


       R1  R2  R3  R4  R5  R6 ?
F1    1    2    3    4   5
F2    2    3    4    5   1
F3    3    4    5    1   2
F4    4    5    1    2   3
F5    5    1    2    3   4
Reply
#2
Hi Anthony,

OK I'll bite,

     r6
f1  1
f2  2
f3  3
f4  4
f5  5

In a modulus world history is recycled.

Why do I think this might be a trick question?
B += x
Reply
#3
Hi bplus

Welcome to the Forum Smile

OK! Well done...

      R1  R2  R3  R4  R5  R6 ?
F1    1    2    3    4    5    1
F2    2    3    4    5    1    2
F3    3    4    5    1    2    3
F4    4    5    1    2    3    4
F5    5    1    2    3    4    5

So now we can Start the interesting stuff like "How can we write a QB64 Program to do the Forecasting...even better ? 

Below is a more difficult version of the above first example... and at the moment I am not sure what the Results for F1 - F5 / R6 are because I just fired out a Random sample.


      R1  R2  R3  R4  R5  R6 ?
F1    3    5    2    1    4
F2    4    2    5    3    1
F3    1    5    4    3    2
F4    2    1    4    3    5
F5    5    1    2    3    4
Reply
#4
Ah! there's more ;-))

The easy solution that has (again) modulus logic, is make r6 the same as r1 and don't worry about keeping a pattern going for both rows and columns, just rows: r6 through r10 same as r1 through r5, repeat again r11...

This would be easier to code if you start with r0, r1, r2, ... r(m-1) when m = modulus, 5 in our case here.

Then r(n) = r(n mod m)
B += x
Reply
#5
Em! Interesting bplus Smile

Explanations on how One thinks or is thinking are always Great But!...

Will you Stick your Neck out and give us what you think the Answers should be to the New Example below (same as above) ?

Please take your time as I may not be able to reply today again ?



      R1  R2  R3  R4  R5  R6 ?
F1    3    5    2    1    4
F2    4    2    5    3    1
F3    1    5    4    3    2
F4    2    1    4    3    5
F5    5    1    2    3    4
Reply
#6
If I were dumb AI, I would repeat my modulus program logic / theory

      R1  R2  R3  R4  R5  R6 ?
F1    3    5    2    1    4      = 3
F2    4    2    5    3    1      = 4
F3    1    5    4    3    2      = 1
F4    2    1    4    3    5      = 2
F5    5    1    2    3    4      = 5

But, I am not dumb (though immortal) AI, just a human.

I will show even more of my thinking:

Wait, you are the chess guy A.R.B. Chess System... pausing /stalling, hmm... am I playing some sort of chess here?

We are in the AI mode / forum section. (I played Chess all through HS, checked out chess club at college was blown away by higher caliber players. Decided at time I best spend my time learning what my parents mostly were paying for. Never got around to proper chess study... discovered programming, along with reading outside English class and journaling (spellcheckers(sp) flag that word all the time, it should be a word!)

I got an "em" like from a teacher who doesn't want to say I am wrong but... hasn't said well done again either hmm...

I really like my dumb AI answer it is easy to code. An Attractor? (sp flags that too) , need to break away from that pull, if I am on wrong track.

Hey ARB, have you heard/read Daniel Kahneman and Amos Tversky? (more pausing / stalling)
I am currently trying to read Michael Lewis "The Undoing Project", unfortunately it has bogged down in discussion of mathematical psychology (ironically I like both subjects). It is due Wed and I have been distracted by the forum, another Attractor.

"Will you stick your neck out..." what are you? a guillotine operator? ;-)) and in Capitals, capitals in the middle of a sentence. BTW, where/when did that expression come from, Stick your neck out, Paris around the time of the French Revolution?

wait look again
"Will you Stick your Neck out and give us what you think the Answers should be to the New Example below (same as above) ?"

Oh! (same as above) I missed that first time, the matrix IS the same.

ARB did you miss my answer? I repeat, same as above.

      R1  R2  R3  R4  R5  R6 ?
F1    3    5    2    1    4      = 3
F2    4    2    5    3    1      = 4
F3    1    5    4    3    2      = 1
F4    2    1    4    3    5      = 2
F5    5    1    2    3    4      = 5

Solution: repackage same answer in more direct solution.

Sorry, I assumed anyone who knows Chess would know modulus and thus might be able to figure out that was my answer.
B += x
Reply
#7
Hi bplus

First of all Great to hear you had an interest in Chess at some time...But I must let you know modulus has nothing much to do with Chess from a Natural Gifted Chess players point of View! Smile
Yes! in the modern world with Computer Analysis Etc. then it has a role to play concerning the Math behind Chess...But again Great Chess players come from all kinds of backgrounds! from the Very Poor to the Very Rich, So a person who learnt to play Chess sitting on the ground because they could not even afford Chairs, is very unlikely to know what modulus is about...

https://www.dawn.com/news/1199229

Meanwhile back on Earth!

Your Answer this time is more than likely Wrong this time! I hope you like the way I use "more than likely Wrong this time!" Smile

Let's have a look at a Clue do you think the R1 (Result) has the same impact trying to Forecast something as a more recent R5 (Result ?

More Cryptic Clues as we Progress! Smile


Anthony.
Reply
#8
I wish I had the free time to soak all of this into my little overwhelmed and maxed out brain! All this looked so much like it would be fun.

Anthony / bplus, why not write a little game demo for this and share it here on the forum? That could get other people into trying it out and possibly teach them something new.

That was an interesting article you linked to. Thanks for sharing that.
Dedicated to empowering computer programming hobbyists, tinkerers, amateurs, and enthusiasts.
profile for Walter Whitman at Stack Overflow, Q&A for professional and enthusiast programmers


Reply
#9
Don't worry Walter Smile a Program for this is coming after some fun! then the Real Fun starts "Who can make the Best & Most accurate Version...regarding Who can Forecast the most Correct from many Examples!
But first we need to Answer the Why's of it all ?


Anthony.
Reply
#10
Anthony,

Sounds good to me!
Dedicated to empowering computer programming hobbyists, tinkerers, amateurs, and enthusiasts.
profile for Walter Whitman at Stack Overflow, Q&A for professional and enthusiast programmers


Reply
#11
OK! Does anyone understand the Table Attached shown below...

[attachment=1152]
Reply
#12
> Walter I would love to offer up some little program but I yet to figure out what ARB is up to!

> Anthony, you have this fish proper hooked!  ;-))

I am going to do some more public thinking but I see you are back at forum so I will let you post more clues if you can't wait for me to answer about...

Well I can answer that now!

Yes R5 is more important in forecasting than R1.

There my neck is stuck out again! ;-))

But I will take a 2nd guess and think publicly because it's fun too, maybe THIS TIME, I will be closer...
B += x
Reply
#13
Anthony, our posts crossed.

Can you provide a context to the numbers, what kind of data or quantities they represent.

Is the 2nd table on the right of first related to the left in 100's ?
B += x
Reply
#14
Yes! the Tables are Related the One Using the 100's is a Score Table  

There are many ways of doing this I am giving a first Example...We are looking for a Best way ?
Reply
#15
Public thinking:

Just the facts mam:
OK we are looking at scores on table on left and some sort of coding of the scores on left in table on right (assumption) and we are looking to make predictions in R6 column.

One should not assume chess players know what modulus is. Irrelevant but good to know. ( I should have said Chess players who are attending a coding forum.) ;-))

"This time" was emphasized by Anthony also as clue.

Probabilities must be involved.

Look again at table.
Oh! 500 = 1, 400 = 2, 300 = 3, 200 = 4, 100 = 5
It is coding of scores, assumption now fact.

Now what about R6? (I need a nap.)
B += x
Reply
#16
First of all the Tables were Learning Examples...

R5 The most Recent should have the Most impact ? but in some cases it only scored 100

OK! R6 Using 1 to 5 Only Once Each!

Example you can't use two with a Value of 500 

Top down Scores...

First Table 

F1 = 1 = (Score 500)
F2 = 2 = (Score 400)
F3 = 3 = (Score 300)
F4 = 4 = (Score 200)
F5 = 5 = (Score 100)

By Chance the Second Table Scored the Same as Table 1  Smile
Just in a different layout.

Top down Scores...

F1 = 1 = (Score 500)
F2 = 2 = (Score 400)
F3 = 3 = (Score 300)
F4 = 4 = (Score 200)
F5 = 5 = (Score 100)
Reply
#17
When using a Group of Numbers like 1 to 5 (Only once each in any order) 

How can we Analyse them ? to be able to make a Forecast for the next five numbers in Order ?

1 + 2 + 3 + 4 + 5 = Total 15  Average =  3

2 + 5 + 1 + 3 + 4 = Total 15  Average =  3

4 + 3 + 2 + 5 + 1 = Total 15  Average =  3

Attached Below is One way! is it the Best ? is there another Way ?

[attachment=1157]
Reply
#18
Dang! I thought I had an idea but then I see every row AND every column use all 5 digits.

I can't imagine some real situation that that typically happens. Need concrete example.
B += x
Reply
#19
There are many Concrete examples! you could look at the five Results R1 - R5 as being finish positions from a Race something like Horse Racing, or Greyhound Racing with a Dog short! because normally there are Six Dogs in those Races.

http://www.attheraces.com/racecard/Ascot...-2017/1400

https://www.sportinglife.com/greyhounds/...card/21735
Reply
#20
That's how I was trying to see it.

I was imaging Chess players, 5  F1-F5 going 5 rounds R1-R5 and the problem: predict how the 6th round would go.

Of course even in stalemates, judges can decide who played a better game so ranking players 1 to 5 in each round is fine, I can accept that.

But my imaging collapses when each player must go through the numbers 1 thru 5 no repeats, until (I assume) the 6th round prediction.

This, in my opinion, is totally unrealistic and my mind ceases to play with this problem because any practical application of this is no where in (my) sight. Which is why bulk and I ask for concrete example.

You offer a dog race, I see it as the same problem, it is not a real problem because this is not how it works in real world.

There are better dogs that win most rounds and worse dogs that loose most rounds and it would be a  freak accident that every dog would come in 1st, 2nd, 3rd, 4th, 5th over 5 rounds with 5 dogs.

It's true I've never attended a dog race, if this is what happens at them, I better go see one, so I can believe it!  ;-))

PS What my mind will gladly play with (and has) is a Sudoku like puzzle, true it's totally artificial but it's about logic not probability.
B += x
Reply