Breaking Bug - Machine Learning Repository

Introduction

This repository contains the backend code for the Breaking Bug event. The event is organized by IEEE Computer Society, Manipal University Jaipur.

Breaking Bug is an electrifying virtual showdown for tech enthusiasts and coding maestros! An exciting and challenging event where participants step into the shoes of skilled developers and problem-solvers! In this unique competition, their mission is to identify and fix bugs in a GitHub repository across three diverse domains: Frontend, Backend, and Machine Learning (ML).

Pre-requisites

Here are the pre-requisites in point form:

Python libraries:

pandas numpy matplotlib seaborn plotly scikit-learn (metrics, preprocessing, model selection, ensemble) xgboost scipy

My Project Information

Model	Cross-Validation Accuracy	Test Accuracy
Logistic Regression	0.5169	0.4565
Gradient Boosting	0.6329	0.5543
KNeighbors Classifier	0.5785	0.5435
Decision Tree Classifier	0.6026	0.5435
Random Forest Classifier	0.6498	0.5652
AdaBoost Classifier	0.5652	0.4783
XGBoost Classifier	0.6352	0.5978
Support Vector Machine	0.5942	0.5000
Naive Bayes Classifier	0.3696	0.3043

Best Model: XGBClassifier Best Model Cross-Validation Accuracy: 0.3696 Best Model Test Accuracy: 0.5978

Given Project Information (Breaking-Bug):

Project Information

Here’s a revised summary focusing on the ML-related details:

Points Distribution

The maximum attainable points for this project are 1000. The points are distributed as follows:

Difficulty Level	Points
Very easy	20
Easy	30
Medium	40
Hard	75
Easter egg	100
Total	1000

Here are the columns from the dataset, with their descriptions:

Dataset Columns

• id: Unique ID
• age: Age in years
• sex: Gender
• dataset: Location of data collection
• cp: Chest pain type
• trestbps: Resting blood pressure
• chol: Cholesterol measure
• fbs: Fasting blood sugar
• restecg: ECG observation at resting condition
• thalch: Maximum heart rate achieved
• exang: Exercise induced angina
• oldpeak: ST depression induced by exercise relative to rest
• slope: The slope of the peak exercise ST segment
• ca: Number of major vessels (0-3) colored by fluoroscopy
• thal: Thalassemia
• num: Target [0 = no heart disease; 1, 2, 3, 4 = stages of heart disease]

Here’s the updated table without the "Best Hyperparameters" column:

Model Performance

Model	Cross-Validation Accuracy	Test Accuracy
Logistic Regression	0.5115	0.5109
Gradient Boosting	0.6396	0.5978
KNeighbors Classifier	0.5767	0.5870
Decision Tree Classifier	0.5840	0.5761
AdaBoost Classifier	0.6058	0.5978
Random Forest	0.6288	0.6739
XGBoost Classifier	0.6263	0.6413
Support Vector Machine	0.5877	0.5870
Naive Bayes Classifier	0.5780	0.5435

Best Model: XGBoost Classifier
Best Model Cross-Validation Accuracy: 0.6263
Best Model Test Accuracy: 0.6413

Made by IEEE Computer Society- Manipal University Jaipur