0
$\begingroup$

I have a ranked list of rows of 100 lines of data

1- 8 4 0 5 9 3

2- 0 3 3 5 3 2

3- 0 0 2 4 0 2

..

100- 0 2 3 2 2 0

Is it possible to predict

a) when given a new sequence where it would fit?

b) a way to generate a row that would fit into the top 1, top 5, top 10

$\endgroup$

closed as unclear what you're asking by Stephen Rauch, Rejaul Karim, Toros91, TitoOrt, Sean Owen Dec 15 '18 at 5:09

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

0
$\begingroup$

The short answer is yes.

What you are looking at would be developing a model that would regress an input of 6 integers to a single integer between 0 and 100. $f(x) : \mathbb{Z}^6 \rightarrow \mathbb{Z}^1$

This could also be done as as 100 class classification problem.

The regression approach would be better in my opinion.

One big question that you need to know to better determine if machine learning is the correct approach: are the sequences (e.g., 0 3 3 5 3 2) the only sequences that = the given ranks (e.g.,1). If so you only need to match sequences, no need for M L.

If no, and you chose to develop an ML approach, an issue you have is that you only have 1 observation for each integer.

Depending on the variability of the input integers and their output rank {1 - 100} you would want multiple observations corresponding to each integer.

One last thing, I would start with one linear and one non-linear algorithm to see which method fits this data best.

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.