Learn Data Skills
Beta
Zakaria BOULKHIR

Zakaria BOULKHIR

Certified

Data scientist

Reply Inc. | Brussels

Technologies

My Portfolio Highlights

My New Workbook

Experiment with the AI assistant

My New Course

Introduction to Python

My New Workbook

Experiment with the AI assistant

My New Workbook

Experiment with the AI assistant

Data wizard, conjuring insights from the depths of complex datasets.

My Work

Take a look at my latest work.

DataLab

Experiment with the AI assistant

DataLab

Experiment with the AI assistant

DataLab

Experiment with the AI assistant

DataLab

Experiment with the AI assistant

My Certifications

These are the industry credentials that I’ve earned.

Data Scientist Professional

Data Scientist Professional

DataCamp Course Completion

Take a look at all the courses I’ve completed on DataCamp.

Courses

Introduction to Python
Data Manipulation with pandas

My Work Experience

Where I've interned and worked during my career.

Avantage Reply

Data Scientist

N/A

Leroy Merlin

Data scientist

• As a retail company, Leroy Merlin uses Recommender Systems (SdeR) in many applications: Web, Stores and Preconisation. The goal of the internship relies on providing an exhaustive Bibliography on SdeR. • Provided an implementation of the main State-of-the-art algorithms. Mainly, Matrix Factorisation that takes advantage of its techniques in the Collaborative Filtering framework. • Adapting these algorithms to the retail world and specific uncommon applications such as Preconisation.
Show More

Inria

Statistical analysis and database management

• I studied the problematic of data migration and I worked on the use case of FHIR. • I have acquired more knowledge about the functioning of graph databases during weekly conferences. • I started by statistically analysing the use case which is an important step especially in minimizing the operations’ complexity. A second part focused on the use of a Natural Language Processing (NLP) algorithm along with federated learning.

Inria

Research project : Trajectory clustering

• In a partnership between Inria and Mapwize, I worked on a project that aimed at clustering trajectories. This project is deepened in programming, ML and theoretical mathematics. • The project went through three main steps: re-sampling, optimization and measure definition. Re-sampling consisted on representing data as line segments (LS) instead of points. We have also added another dimension which indicated the direction of each LS. This variation boosted our model to the state-of-the-art algorithm. • Optimization helped reducing dimensions using Functional Principal Component Analysis (FPCA) which selects more representative features. • Finally, I have equipped the model with different distance measures (angular distance, probabilistic distances). The model was Density-based, as it used the distances and probabilities to make the clusters. • I also considered some deep learning (especially neural networks) technics and algorithms, but sadly they were mostly prone to overfitting.

Inria

Computer science Internship - Browser fingerprinting

• Studying how browser fingerprinting can be used to track users on the web. Removing any new methods that help tracking users. • Statistical study of the survey in the amiunique.org site or some other databases. • Resulted in finding another privacy flow allowing the host websites to detect whether users are using incognito mode or not.

My Education

Take a look at my formal education

Centrale Lille
Université de Lille
Master's degree, Data science in MasterUniversité de Lille & Centrale Lille | 2022
Informatique in LicenceUniversité de Lille | 2020

About Me

Zakaria BOULKHIR

A data scientist at Reply. I am enthusiastic for the industrial and research advancements on the field data science. Having many experiences in the field, I have acquired a practical knowledge and many useful practices in Data science.

Powered by

  • Work
  • Certifications
  • Courses
  • Experience
  • Education
  • About Me
  • Create Your Data Portfolio for Free