---
title: "Discussion on \"Building a Decision-Support Framework When No Dataset Exists\" | Hashnode"
description: "Discussion on \"Building a Decision-Support Framework When No Dataset Exists\". A Q-learning and Markov Chain framework built on 13,777 penalty outcomes across 55 NCAA Division I programmes, from a blank slate to a published preprint."
url: "https://hashnode.com/posts/building-a-decision-support-framework-when-no-dataset-exists/6a8e4eb8edd516b00712ad91"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Building a Decision-Support Framework When No Dataset Exists](https://amruthar18.hashnode.dev/building-a-decision-support-framework-when-no-dataset-exists?utm_source=hashnode&utm_medium=feed)

A Q-learning and Markov Chain framework built on 13,777 penalty outcomes across 55 NCAA Division I programmes, from a blank slate to a published preprint.

Aug 26 · 4 min read · amruthar18.hashnode.dev  
Read the article

-   [#machine-learning](https://hashnode.com/tag/machine-learning)
-   [#reinforcement-learning](https://hashnode.com/tag/reinforcement-learning)
-   [#sports-analytics](https://hashnode.com/tag/sports-analytics)
-   [#data-science](https://hashnode.com/tag/data-science)
-   [#research](https://hashnode.com/tag/research)

## Written by

[Amrutha Ravikumar](https://hashnode.com/@amruthar18)

@amruthar18 · Portland, USA

  
[View profile](https://hashnode.com/@amruthar18)

Turning prediction data into product gears

I'm pursuing my Master's in Applied Machine Intelligence at Northeastern, with a first-author preprint on reinforcement learning in NCAA ice hockey, validated across 55 Division I programmes. My background spans business, management, and Human Resources, which shapes how I think about who uses a model and whether they can trust its output. I write here about the parts of a project that don't make it into a resume, the pipeline before the model, the validation step that's easy to skip under a deadline and focused on building toward user-centric solutions.

## Responses 1

Markdown supported · ⌘↵ to post

[Kartik N V J K](https://hashnode.com/@kartik-nvj)

[Aug 26](https://hashnode.com/posts/building-a-decision-support-framework-when-no-dataset-exists/6a8e4eb8edd516b00712ad91/comment/6a8f3494194331e55bc51743)

The leave-one-season-out call is the right instinct. With five seasons the failure mode is exactly what you named: sharp on training data, dead on a season it never saw. I would also slice-check the rare states, since an 8-state space can hide a class showing up only a few times. Did any state end up too sparse to trust?
