Scribd
Upload
6 views

Instructions

The document provides guidelines and a problem statement for building a model to predict the outcome of football matches from past data. It describes the goal is to predict if the home team will win, lose, or draw for matches in the 2017-18 season. The data provided and expectations for model submissions are also outlined.

Uploaded by

rajdeep.jzs
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
6 views

Instructions

The document provides guidelines and a problem statement for building a model to predict the outcome of football matches from past data. It describes the goal is to predict if the home team will win, lose, or draw for matches in the 2017-18 season. The data provided and expectations for model submissions are also outlined.

Uploaded by

rajdeep.jzs
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Guidelines

At Sahaj, we strive to build high-quality software that has strong aesthetics (is
readable and maintainable), has extensive safety net to safeguard quality, handles
errors gracefully and works as expected, without breaking down.

We are looking for people with data science knowledge coupled with pragmatism to
deliver the implementation to business production environments. The data scientist
should understand the domain and build models that have the ability to deal with
the real-life constraints.

Following is a list of things to keep in mind, before you submit your solution, to
ensure that your model focuses on attributes, we are looking for -

● Have you understood all the variables in the data?


● Have you followed best practices to make sure your model is robust?
● Have you thought about how the model would evolve, over a period of time,
in production, when more data is available?
● Have you made an effort to make your code readable and robust?
● Have you thought about the biases that can be there in the data?

Page 1
Problem Statement

Your goal is to build a model to predict the outcome of a football match, given data for the past 9
years. All the football matches from 2009 to 2017 are covered in the dataset.

Based on the dataset provided, your goal should be to come up with an optimal solution to
predict if a Home Team would win or lose or draw a game (column name ​FTR​) for the year of
2017-18.

We are looking to compare multiple approaches and choose the one that performs the best.

Use of external data sources is encouraged, with some recommendation. Of course, the actual
results of these matches can be easily downloaded from the web. ​However, this problem
statement is intended to be for fun and learning. ​You can choose to enrich the dataset by using
other publicly available European football league datasets, e.g. ​http://www.football-data.co.uk/
or ​http://football-data.mx-api.enetscores.com/

The train and the test dataset is not randomly sampled for testing and training

Page 2
About the Data
The data is collected from ​http://www.football-data.co.uk/​ and consists of different leagues.

Column Details

Name Description

HomeTeam Home Team

AwayTeam Away Team

FTR Full-Time Result (H=Home Win, D=Draw, A=Away Win)

HTHG Half Time Home Team Goals

HTAG Half Time Away Team Goals

HS Home Team Shots

AS Away Team Shots

HST Home Team Shots on Target

AST Away Team Shots on Target

AC Away Team Corners

HF Home Team Fouls Committed

AF Away Team Fouls Committed

HC Home Team Corners

HY Home Team Yellow Cards

AY Away Team Yellow Cards

HR Home Team Red Cards

AR Away Team Red Cards

Date On which day the match was played

league Under which league the match was played

Page 3
Expectations from the submission
1. Approach to the dataset.
2. Choice of the model(s).
3. Model validation approach (used locally).
4. Choice of features used for the model creation.

Page 4

You might also like