"I like problems that can be written down, modeled, optimized, and then solved."
I'm Faycal Billel Alikacem, a Master's student in Stochastic Modeling & Forecasting with a background in Operational Research and Applied Mathematics.
My work lives at the intersection of mathematics, algorithms, optimization, and artificial intelligence.
I enjoy going beyond using a model as a black box — understanding the mathematics behind it, formulating the problem, implementing the method, and comparing the results experimentally.
Mathematics
│
├── Probability & Stochastic Processes
├── Statistics & Time Series
└── Optimization & Graph Theory
│
▼
Algorithms
│
├── Exact Methods
├── Heuristics
├── Metaheuristics
└── Learning-based Methods
│
▼
Intelligent Systems
|
Machine Learning · Deep Learning Reinforcement Learning · NLP Transformers · Representation Learning |
Stochastic Processes · Simulation Statistical Modeling · Time Series Forecasting · Monte Carlo Methods |
|
Linear Programming · MILP Combinatorial Optimization Network Optimization · Metaheuristics |
Graph Algorithms · Graph Models Network Design · Routing Graph-based Optimization & Learning |
|
A Python toolkit for stochastic processes, probability models, and simulation.
|
Optimization models for telecommunication network problems, with a focus on mathematical programming.
|
|
A hybrid approach combining reinforcement learning and adaptive large neighborhood search for vehicle routing.
|
Exploring forecasting by combining market time series with textual information.
|
|
Breaking down Transformer architectures by implementing the underlying mechanisms from the ground up.
|
Exploring optimal transport as a mathematical tool for domain adaptation in machine learning.
|
▸ Optimization for real-world network problems
▸ Mathematical programming & solver-based methods
▸ Time-series modeling & forecasting
▸ Graph algorithms & graph-based learning
▸ Deep learning architectures
▸ Reinforcement learning
▸ AI + Optimization for decision making
▸ Research-oriented implementations
I like projects where there is a genuine problem behind the code.
Problem
↓
Mathematical formulation
↓
Baseline / exact model
↓
Algorithm design
↓
Implementation
↓
Experiments
↓
Comparison & analysis
↓
Reproducible results
The objective is not simply to make something run.
It is to understand why it works, when it fails, and how it can be improved.
