AI/ML Roadmap

Vision Statement

My objective is to build research-grade expertise in efficient deep learning and computer vision, grounded in an Electrical and Electronics Engineering foundation. The focus is on model compression, object detection architectures, and applied AI research — producing work that is both theoretically rigorous and deployment-ready.

Progress is measured through production-grade projects, ablation studies, and peer-reviewed quality technical deep dives.


Phase I: Foundational Mastery ✓

Goal: Solidify core ML principles and the mathematical foundations of deep learning.

Focus AreaCore ConceptsKey Output
Applied MathematicsLinear algebra for DL, multivariable calculus, optimization theoryDeep Dive: The Calculus of Loss Functions and Backpropagation
Core ML & StatisticsSupervised/unsupervised learning, probability, Bayesian methodsIBM AI Engineering Professional Certificate (Coursera)
AlgorithmsComplexity, data structures, dynamic programmingStanford Algorithms Specialization (Coursera)
Python & ToolsNumPy, PyTorch basics, GitVersion-controlled project repositories

Phase II: Deep Learning Specialization ✓ ✓ 🔄

Goal: Master CV architectures and frameworks required for research-level work.

Computer Vision

Focus AreaArchitectures & TechniquesKey Output
CNN MasteryConvolutional layers, transfer learning (ResNet, VGG)Deep Dive: Weight Initialization Schemes in CNNs
Object DetectionYOLO, R-CNN, RT-DETR, DETRProject: Real-Time Object Detection (GitHub)
Vision TransformersViT, self-attention, cross-attentionOngoing — RT-DETR architecture study

Systems & Acceleration

Focus AreaTechniquesKey Output
CUDA ProgrammingParallel computing, memory management, kernel optimizationNVIDIA: Getting Started with Accelerated Computing in CUDA C/C++
Embedded AIReal-time inference, edge constraintsSignal Processing Toolbox (C++ / Qt)

Phase III: Applied AI Research 🔬 Current

Goal: Contribute original work at the intersection of model efficiency and computer vision.

Focus AreaResearch DirectionKey Output
Knowledge DistillationLogit-KD vs Feature-KD for object detectors, teacher-student dynamicsCS229 Project: RT-DETR Distillation — 12-config ablation on COCO
Model CompressionPruning, quantization, architecture searchOngoing
Representation LearningFeature alignment, intermediate supervisionCS229 theoretical analysis
Benchmarking & MLOpsFPS benchmarking, VRAM optimization, reproducibilityRTX 3050 → Colab → Kaggle GPU pipeline

Phase IV: Research Publication & Positioning 📄 Next

Goal: Translate applied research into publishable work and establish a research identity.

Focus AreaDirectionKey Output
Technical WritingConference-style papers, ablation reportingCS229 final report → arXiv pre-print
Open SourceWell-documented repos, reproducible baselinesPublic RT-DETR distillation codebase
Research PositioningApplied AI Scientist roles, research engineer trackPortfolio: umutonuryasar.com + GitHub

Last updated: March 2026