defter / katalog / EEE 443

EEE 443

Neural Networks

EEE 443 builds your understanding of deep learning from the ground up, starting from how a single neuron computes, then layering on backprop, recurrence, convolution, and attention until you can reason about why modern architectures like transformers and diffusion models actually work. Most of your grade comes from four projects where you implement and train these networks yourself, so the math from linear algebra and probability stops being abstract and becomes the thing you debug at 2am. It sits as the natural follow-up to a signals/ML-flavored background and is the course that prepares you for research or industry work in computer vision, NLP, and generative modeling.

Kredi3ECTS5FakülteFaculty of EngineeringBölümElectrical and Electronics EngineeringKoordinatörTolga Çukur

Ön koşulEEE 321 veya EEE 391

→ STARS müfredatı (resmi syllabus)

Haftalık müfredat 14 hafta

Hafta 114–20 Eyl

Machine learning matematiği ve neural network girişi

Primer on Linear Algebra, Probability and Optimization for Machine Learning / Introduction to Neural Networks (NNs)

linear algebraprobabilityoptimizationneural networks

Hafta 221–27 Eyl

Nöronların biyofiziksel ve matematiksel modelleri

Neurons: Biophysical and Mathematical Models

neuronsbiophysical modelsmathematical models

Hafta 328 Eyl – 4 Eki

Neural network yapıları: perceptron

Neural Network Structures: Perceptrons

neural network structuresperceptron

Hafta 45–11 Eki

Öğrenme algoritmaları ve türleri

Learning Algorithms: Supervised, Unsupervised, Reinforcement Learning

supervised learningunsupervised learningreinforcement learning

Hafta 512–18 Eki

Tek katmanlı NN eğitimi

Training Single Layer NNs

single layer NNtraining

Hafta 619–25 Eki

Çok katmanlı NN eğitimi ve back propagation

Training Multilayer NNs: Back Propagation, Empirical Risk Minimization

multilayer NNback propagationempirical risk minimization

Hafta 726 Eki – 1 Kas

Optimization yöntemleri ve generalization

Optimization Methods and Generalization

optimization methodsgeneralization

Hafta 82–8 Kas

Recurrent neural network'ler: BPTT ve LSTM

Recurrent Neural Networks (RNNs): BP Through Time, LSTM

RNNbackpropagation through timeLSTM

Hafta 99–15 Kas

Restricted Boltzmann Machines ve Contrastive Divergence

Restricted Boltzmann Machines (RBMs): Contrastive Divergence

Restricted Boltzmann MachinesRBMContrastive Divergence

Hafta 1016–22 Kas

Autoencoder türleri ve stacked ağlar

Autoencoders: Denoising, Contractive, Stacked NNs

denoising autoencodercontractive autoencoderstacked neural networks

Hafta 1123–29 Kas

Deep learning: autoencoder ve belief networks

Deep Learning: Deep Autoencoders, Deep Belief Networks

deep learningdeep autoencodersdeep belief networks

Hafta 1230 Kas – 6 Ara

Computer vision: convolutional ve transformer ağlar

Vision Models: Convolutional and Transformer Networks for Computer Vision

convolutional networktransformer networkcomputer vision

Hafta 137–13 Ara

Dil modelleri: RNN ve Transformer

Language Models: RNNs and Transformers for Natural Language Processing

language modelRNNTransformernatural language processing

Hafta 1414–20 Ara

Generative modeling yaklaşımları

Generative Modeling: Adversarial Models, Variational Models, Diffusion Probabilistic Models

adversarial modelsvariational modelsdiffusion probabilistic models

Değerlendirme 100% · 6 adım

25%

15%

10%

30%

Quiz Quiz Average 25%

Homework Tutorial Average 15%

Project 4×Project Assignment 1 60%

en büyük tek kalem %30 · sınav ağırlığı %25 · 22 dönem ortalaması 2.79 (1088 öğrenci) nasıl hesaplanıyor

Önerilen kaynaklar 3 kitap

📕

Zorunlu

Deep Learning

Ian Goodfellow and Yoshua Bengio and Aaron Courville

2016 · The MIT Press

📖

Önerilen

Neural Networks and Learning Machines

Simon S. Haykin

2009 · Upper Saddle River: Pearson Education

📖

Önerilen

Introduction to Artificial Neural Systems

J. M. Zurada

1992. · West Pub. Co

Bu dersi alınca · 4 öğrenme çıktısı

Understand the basic neuron structure, related mathematical neuron models and basic neural network models.Quiz Average, Project Assignment 1
Understand the basisc of learning in neural systems, formulate various learning rules and investigate their propertiesQuiz Average, Project Assignment 1, Project Assignment 2, Final Project
Understand the relation between dynamical systems and some neural network structures, utilize feedback control systems ideas to investigate their propertiesFinal Project
Apply the neural network structures in solving some engineering problemsProject Assignment 2, Project Assignment 3, Final Project

Bilkent'in resmî syllabus'ünden. Sağdaki etiket o çıktının hangi değerlendirmeyle ölçüldüğünü söylüyor.

🤖 GenAI politikası

Generative AI tools (e.g., large language models, coding assistants, and image-generation tools) may be used for take-home assessments (i.e., tutorial/homework and project assignments) to assist with coding, and improving writing. Any use of generative AI must be clearly disclosed, including the name/version of the tool and a brief description of how it was used. Upon request, students should be ready to provide prompts, outputs, and execution logs pertaining to their genAI use for all assignmen

Ders notları · henüz yok

EEE 443 için defter ekibi henüz not yazmadı.

İlk dosyayı sen atarsan: not, slayt, geçmiş sınav, çözüm, cheat-sheet, ne varsa. defter ekibi öğrenci paylaşımlarından bu dersin notlarını yazar. Drive linki / PDF / ZIP, hepsi olur.

← katalog

Geçmiş GPA dağılımı 22 dönem · ort. 2.79

Dönem	Course CPA
2025-2026 Fall	3.25	1 sec · 67 öğr
2024-2025 Fall	3.34	1 sec · 65 öğr
2024-2025 Spring	3.03	1 sec · 64 öğr
2023-2024 Fall	3.37	1 sec · 46 öğr
2022-2023 Fall	2.67	1 sec · 59 öğr
2022-2023 Spring	3.06	1 sec · 30 öğr
2021-2022 Fall	3.24	1 sec · 53 öğr
2021-2022 Spring	2.69	1 sec · 38 öğr
2020-2021 Fall	2.87	1 sec · 50 öğr
2019-2020 Fall	2.86	1 sec · 71 öğr

Aggregate course GPA · Bilkent STARS'tan public data. Hoca-bazlı per-section detayı için STARS evaluation report →. Öğrenci anket cevapları KVKK kapsamında defter'de tutulmaz. Tüm derslerin ortalamaları →

Bu dönem · 2026-2027 Güz · 1 şube · 60 kontenjan

haftada 3 saat ders

Haftalık ızgarada 4 saat görünüyor, kayıt sistemi dersi 3 saat yazıyor. Aradaki 1 saat yedek saat: Bilkent programlarında derse genelde bir fazla saat ayrılır ve çoğu zaman kullanılmaz. Hangisinin düşeceği dönem başlayınca belli olur.

Ömer Morgül

Çar08:30–10:20Cum13:30–15:20

kişilik

programıma ekle → CGPA'ya ekle kayıt tarihleri

Genelde sadece güzSon yıllarda 21 kez güz döneminde açılmış, 6 kez bahar. Bu dönem alamazsan bir sonraki güzü beklemen gerekebilir.

⚠️ FZ engelleyen şartlar

All in-class and take-home assignments should be completed and submitted. Grades, except the final exam, should justify a letter grade of D or better. (Note that meeting this criterion does not guarantee that you will eventually pass the course. Your letter grade still depends on your overall performance including the final.)

Hocalar 1 bu dönem · 3 geçmiş

Bu dönem (2026-2027 Fall) · 1 section

Ömer Morgül

Geçmişte ders veren (3 kişi)

Erdem Koyuncu, Tolga Çukur, Cem Tekin