This Git repository contains a learning trace of my journey in machine learning (ML) and deep learning (DL). The purpose of this repository is to document my progress, keep track of the resources I have used.
The repository contains the following folder:
-
01-foundations: in this directory, you will find Jupyter notebooks designed to facilitate my learning of essential Python libraries, including Numpy, Pandas, Scipy, Matplotlib, as well as ML and DL frameworks like Scikit-learn and Pytorch.
-
02-paper-implementations: using well-known Python frameworks and libraries, the most renowned scientific papers implemented in this folder. Papers are about computer vision, natural language processing (NLP), speech recognition, and reinforcement learning.
-
03-building-makemore (Andrej Karpathy): this folder contains the Jupyter notebooks related to the Karpathy's cours about makemore
-
04-building-micrograd (Andrej Karpathy): this folder contains the Jupyter notebooks related to the Karpathy's cours about micrograd
-
04-from-scratch: this folder contains Jupyter notebooks that implement some DL & ML algorithms such as kmeans, logistic regression, decision tree, SVM, CNN & RNN from scratch in order fully understand this algorithm.
-
05-data-preprocessing: this folder contains notes & notebooks about various data processing tasks, such as how to handle missing data, how to generating synthetic data (for exemple the GAN algorithm), how to join datasets, and how to perform text and video processing.
-
06-ML-for-production : this folder contains notes & notebook related to course about deployment of ML project in production. For exemple the coursera specializations: ML engineering for production (mlops). It's can be found here
Tools and frameworks encountered during my journey:
- huggingface: Similar to GitHub, there is a platform designed specifically for storing machine learning models, datasets, and applications (referred to as "spaces"). This platform also provides libraries to facilitate the integration of their models and datasets into your code
- DVC : Open-source, Git-based data science. Apply version control to machine learning development, make your repo the backbone of your project, and instill best practices across your team.
- joblib: Joblib is part of the SciPy ecosystem and provides utilities for pipelining Python jobs. It provides utilities for saving and loading Python objects that make use of NumPy data structures, efficiently.
Source encountered during my journey:
- MIT Press book Ian Goodfellow: the most comprehensive book available on the deep learning and available as free html book.
- Andrew Ng: Co-Founder of Coursera, and an Adjunct Professor at Stanford University.
- andrej karpathy: IA God. author of the series of videos NN zero to hero.