🚀 75 Days Kaggle Challenge – ML from Scratch

📌 Introduction

Hey everyone! Myself Pathik Viramgama (Pr0fess0r), and I am on a 75-day Kaggle challenge where I will dedicate 300 hours (4 hours per day) to rise as high as possible in the Kaggle rankings. 🚀

This GitHub Repository serves as a structured log of my journey, where I will be implementing machine learning (ML) from scratch, moving from basic concepts to deep learning. It will be regularly updated with Kaggle Notebooks that I create during the challenge.

---

📚 Table of Contents

1️⃣ Datasets

All datasets used in this challenge are stored in the /Data directory.

Dataset Name	Description
banana_quality_dataset.csv	Dataset related to banana quality classification
heart_dataset.csv	Dataset related to heart disease classification
Salary_dataset.csv	Dataset used for salary prediction (Linear Regression)

---

2️⃣ Notebooks

All ML implementations are in the /Notebooks directory.

Notebook Name	Description
n-4-oops-in-python.ipynb	Object-Oriented Programming (OOP) in Python
n1-linear-regression-scratch-with-ols.ipynb	Linear Regression from Scratch using Ordinary Least Squares (OLS)
n2-linear-regression-scratch-with-gd.ipynb	Linear Regression from Scratch using Gradient Descent (GD)
n3-logistic-regression-scratch-binary.ipynb	Logistic Regression for Binary Classification from Scratch
n3-logistic-regression-scratch.ipynb	Logistic Regression from Scratch using Gradient Descent (GD)
n4-knn-scratch-2-cat-euclidian-distance.ipynb	K-Nearest Neighbors (KNN) from Scratch using Euclidean Distance for Binary Classification
n5-naive-bayes-scratch-gaussian.ipynb	Naive Bayes Classifier from Scratch using Gaussian Distribution
n6-decision-tree-classifier-scratch-2-cat-gini.ipynb	Decision Tree Classifier from Scratch for Binary Classification using Gini Impurity
n7-random-forest-scratch-2-cat-majority.ipynb	Random Forest Classifier from Scratch for Binary Classification using Majority Voting
n8-linear-svm-scratch-2-cat.ipynb	Support Vector Machine (SVM) from Scratch for Binary Classification using Linear Kernel
n9-k-means-clustering-scratch.ipynb	K-Means Clustering from Scratch

---

🔥 Goals & Progress

✅ Challenge Breakdown

• Total Duration: 75 Days
• Total Hours: 300 Hours (4 hours/day)
• Objective: Climb the Kaggle rankings by solving real-world ML problems.

---

📌 What This Repository Covers

✔️ Core ML Concepts from Scratch
✔️ Mathematical Derivations of Algorithms
✔️ Optimized Implementations
✔️ Moving towards Deep Learning

---

📌 How to Use This Repository

1. Clone the Repository

   git clone https://github.com/Pr0fess0rOP/ML-Scratch_Coding.git
   cd ML-Scratch_Coding

---

📊 Roadmap

This repository will be progressively updated. Below is the expected difficulty curve:

📌 Phase 1 - Basics

• Linear Regression (GD, OLS)
• Logistic Regression
• K-Nearest Neighbors (KNN)
• Naive Bayes Classifier

📌 Phase 2 - Advanced ML

• Support Vector Machines (SVM)
• Decision Trees
• Random Forests
• K-Means Clustering

📌 Phase 3 - Deep Learning

• Artificial Neural Networks (ANN)
• Convolutional Neural Networks (CNN)
• Recurrent Neural Networks (RNN)

📌 Phase 4 - State of the Art

• Stable Diffusion
• Transformers (BERT, LSTMs)
• Audio Classification

---

🤝 Contributions

I welcome contributions to improve this repository! You can contribute in the following ways:

---

🔹 Report Issues (Not functional)

• Found a bug or error? Open an Issue with details.
• Suggest an improvement or feature request.