index

About me:

Hey there! , My name is Dhia and I’m an aspiring Software Engineer/ML Engineer and a forever student, in a journey of self-learning assisted through Holberton School, edx, and coursera.

Contacts:

• CV: Dhia_Dahmeni_cv.pdf
• Email: dhiadah@gmail.com
• Github: cryptolake
• LinkedIn: Dhia Dahmeni

Projects:

• Car Detector v2.0
  • A full rework of Car detector v1.0
  • Car extraction model based on Yolos enabling multiple cars in a picture and car/no car verification
  • A custom model based on pre-trained Xresnet with tuned hyper-parameters
  • Advanced Data Cleaning
  • Asynchronous API using FastApi
  • Full custom inference pipeline
• Car Detector
  • A car brand/model/year/color classifier
  • A custom model based on pre-trained ResNet
  • More than 20k+ images of cars scraped and labeled
  • Full custom inference pipeline
  • Deployment of model using docker on bare metal
• Time series forecasting
  • Full Custom RNN-LSTM model for predicting Bitcoin prices
  • EDA walk-through and gathering insights from the provided data
• Machine Learning from scratch
  • This is more than a year of various projects
  • Applied Math
  • Classical neural networks from scratch (only numpy) trained on MNIST
  • NLP from scratch RNN/LSTM as well as transformers (thanks karpathy)
  • Convolutions networks and GANs
  • Training bots to play classical games (Breakout, Tetris…)
  • Unsupervised Learning methods like clustering and dimensionality reduction
  • Advanced pandas/SQL
• QA BOT
  • QA bot based on on a Bert
  • Semantic search using the FAISS index
• Breakout AI
  • RL agent trained on custom deep Q network
  • Implementing Deepmind Q-learning paper
• Simple Shell
  • A simple Linux shell clone of /bin/sh without using standard lib
  • Implementing cd, env, and exit as custom commands
• Monty
  • Implementing Monty byte-code interpreter
  • Custom Queue implementation
• Custom Printf
  • Custom Printf function from scratch
• Creative Automation System
  • Ad generator using AI
  • Dynamic parameters like brand name, product name, and description
  • Ad Image generation and templates
• Airbnb Clone
  • A full-stack Airbnb clone
  • Base models for data and custom admin console
  • Multiple storage implementation: JSON or MySQL DB
  • Unit testing
• Interview/DSA preparation
  • Some Leetcode/interesting problems to solve
• Applied Fun Math
  • Some fun applied math
  • Creating pure c visualizations of the mandelbrot and julia sets
• Project Euler
  • my project euler answers
  • (getting back to this)

Machine learning:

• Hyperparameter Tuning With Bayesian Optimization: Hyperparameter Tuning

• Time series forecasting: My experiment forecasting bitcoin

• Activation: Activation Functions

• Optimization: Optimization methods

• Regularization in ML: Regularization techniques

• Summary of the AlexNet paper: My summary of the AlexNet paper

• My transfer learning experiment: Transfer learning experiment

• computer vision: My computer vision notes

• Orthogonalization

C programming:

• C compilation under the hood

• C static libraries

• Hard links vs Symbolic links, What’s the difference?

• Static Libraries vs Dynamic Libraries in C

• What Happens When you Type “ls -l *.c” in the shell

Mathematics:

• Unit 1 Probability Models and Axioms

• Unit 4 Discrete random variables

• Sets, sequences and series

What is this:

I decided to use obsidian as my knowledge base for studying machine learning/Data Science, software engineering, and random subjects So I will be touching on multiple subjects throughout my journey and will be publishing all of my articles here! Welcome!