Multi-Class Prediction of Obesity Risk- Kaggle Dataset

Introduction

The goal of this competition is to utilize various factors to predict obesity risk in individuals, which is related to cardiovascular disease. On-time prediction using Machine Learning algorithms helps in identifying factors responsible for obesity and cardiovascular diseases like means of transport used, age, gender, weight , height and others.

Data Distribution for various obesity class values per gender is shown above. Data distribution helps in understanding data and data analysis, feature engineering and machine learning algorithms helps in identifying factors for obesity in males and females.

Data Source

The Data Source for Multi-Class Prediction of Obesity Risk is taken from Kaggle. Click here for the dataset. 

In the data source, we have three CSV files:

train.csv – the training dataset; ‘NObeyesdad’ is the categorical target.

test.csv – the test dataset; your objective is to predict the class value of ‘NObeyesdad’.

sample_submission.csv – a sample submission file in the correct format.

Objective

The objective of this competition is to analyse various features to predict obesity risk in individuals, which can eventually lead to cardiovascular disease. In training data, “NObeyesdad’’ is a categorical target with ‘Overweight_Level_II’,  ‘Normal_Weight’,  ‘Insufficient_Weight’, ‘Obesity_Type_III’, ‘Obesity_Type_II’, ‘Overweight_Level_I’, ‘Obesity_Type_I’ as possible values. The goal is to predict the probability of various classes of “NObeyesdad’, the target variable for test data, and submit it in ‘sample_submission.csv’ format.

Evaluation Metrics

Submissions are evaluated on the accuracy score between the predicted probability and the observed target.

Kaggle Notebook

To learn more about data visualization, feature engineering, various machine learning algorithm implementations, and final results, kindly go to the Kaggle notebook.

https://www.kaggle.com/code/playingmyway/multi-class-prediction-of-obesity-risk

Stay Tuned!!

Learn learn to predict an introvert from an extrovert on a Kaggle dataset in detail by clicking on the link below:

Predict Introverts and Extroverts on Kaggle Dataset

Keep learning and keep implementing!!