Machine-Learning-from-Scratch

Machine Learning from Scratch

-- Project Status: [Work in progresss]

Project Intro/Objective

The purpose of this project is to build many (if not all) of machine learning tools from scratch. Many of the codes will take inspiration from textbooks (primarily books by Chollet, Géron and Grus) as well as scipy, numpy, pandas and scikit-learn. There will be two main sections: the first will focus on non-neural-network machine learning while the second will be on neural-network (and further on deep-learning). I would like this project to serve as introduction and detailed analysis of inside-the-box implementations of many of the packages in machine learning.

Methods Used

Data scrapping, data cleaning and data feature engineering
Machine Learning: linear and logistic regresssion, k-Nearest-Neighbors (kNN), Naive Bayes, decision trees and random forest, clustering, Natural Language Processing (NLP), recommender systems (filterings)

Technologies

Python
Numpy,Pandas, Scikit-learn
Jupyter

Outline [Tentatively based on Grus' book]

Chapter 1: Introduction. Some examples to get started
Chapter 2: Python Crash Course. I didn't put anything here since they are standard Python materials.
Chapter 3: Visualizing Data
Chapter 4: Linear Algebra
Chapter 5: Statistics
Chapter 6: Probability
Chapter 7: Hypothesis and Inference
Chapter 8: Gradient Descent
Chapter 9: Getting Data
Chapter 10: Working with Data
Chapter 11: Machine Learning
Chapter 12: k-Nearest Neighbors
Chapter 13: Naive Bayes
Chapter 14: Simple Linear Regression
Chapter 15: Multiple Regression
Chapter 16: Logistic Regression
Chapter 17: Decision Tree
Chapter 18: Neural Networks
Chapter 19: Clustering
Chapter 20: Natural Language Processing
Chapter 21: Neutwork Analytic
Chapter 22: Recommender System
Chapter 23: Databases and SQL
Chapter 24: MapReduce

References

Data Science from Scratch (Grus)
Hands-on Machine Learning with Scikit-Learn, Keras and TensorFlow (Géron)
Deep Learning with Python (Chollet)