К содержимому
learnspaceYOUR NEXT CHAPTER
ПРОСТРАНСТВО ОБУЧЕНИЯ
ГлавнаяКаталог курсовМоё обучениеCoursera

Знания без границ

Учитесь у лучших университетов и компаний мира.

Открыть Coursera
Интеграция
Пространство университета
Моё пространствоСтраница курса
↵
ЯЛичный кабинетСтудент
© 2026 LearnSpaceКаждый день — возможность узнать больше.Помощь
Deep Learning & Modern Architectures · LearnSpace
Назад в каталог
courseraАнализ данных

Deep Learning & Modern Architectures

Курс от Coursera
Уровень не указан≈ 25.5 чАнглийский
О курсеНавыкиПрограммаПреподаватели

О курсе

In this course, you will learn how to design, train, optimize, and troubleshoot modern deep learning models used across today’s most advanced AI systems. You will build feedforward neural networks, convolutional neural networks for computer vision, recurrent and LSTM models for sequential data, and transformer-based architectures for natural language processing. You will also explore generative models such as autoencoders, VAEs, and GANs, and learn how to apply transfer learning with pre-trained models to accelerate development. By completing this course, you will gain the practical skills needed to move from foundational neural networks to state-of-the-art architectures that power real-world applications in vision, language, and generative AI. You will develop intuition for optimizing model performance, addressing training instability, and managing overfitting, while gaining hands-on experience implementing models in modern deep learning frameworks. What makes this course unique is its end-to-end architectural progression—guiding you from core concepts through CNNs, RNNs, transformers, and generative models in a single, cohesive learning path. Whether you are preparing for advanced AI roles or strengthening your deep learning foundation, this course equips you with the skills and confidence to build and deploy modern neural network solutions.

Навыки, которые вы освоите

Artificial Neural NetworksModel TrainingFine-tuningGenerative Model ArchitecturesGenerative Adversarial Networks (GANs)AutoencodersNetwork ArchitectureConvolutional Neural NetworksLarge Language ModelingGenerative AITransfer LearningDeep LearningRecurrent Neural Networks (RNNs)Model OptimizationComputer Vision

Программа курса

8 модулей · 99 учебных материалов

01Start Here: Get Oriented and Check Your Skills2 материалов
Start Here: How This Skill-Based Course WorksЧтениеSkill Diagnostic: Find Your Recommended Starting PointЗадание
02Introduction to Deep Learning & Neural Networks23 материалов

Deep Learning Foundations

How to use Jupyter NotebookЧтение

Учитесь у экспертов

Professionals from the Industry

Преподаватель курса

Deep Learning & Modern Architectures
В каталоге вашей программы

Инвестируйте в себя

Новые знания — в удобное для вас время.

Начать на Coursera

Обучение откроется на Coursera
в новой вкладке

Обучение на Coursera

≈ 25.5 ч

8 модулей

Язык: Английский

Субтитры: Арабский, Французский, Узбекский, Украинский, Китайский (Китай), Греческий, Итальянский, Бразильский португальский, Вьетнамский, Нидерландский, Корейский, Немецкий, Пушту, Русский, Тайский, Индонезийский, Шведский, Турецкий, Азербайджанский, Испанский, Хинди, Японский, Казахский, Венгерский, Польский

Часть программы вашего университета
What Is Deep Learning and How Do Neural Networks Work?Чтение
Building a Neural Network and Visualizing the Forward PassВидео
Visualizing the Backward Pass and Gradient Flow with AutogradВидео
Lab - Build and Visualize a Perceptron from ScratchЛабораторная

Building a Single-Layer Neural Network

Building the Perceptron Forward Pass in PyTorchВидеоThe Perceptron Learning Rule and Weight UpdatesЧтениеTraining the Perceptron with the Perceptron Learning RuleВидеоLab - Build Your Own Perceptron for Binary ClassificationЛабораторная

Working with PyTorch

What Are Tensors and Why They MatterЧтениеTensor Operations and Best PracticesЧтениеGetting Started with Tensors in PyTorchВидеоReshaping Tensors and Using GPUs in PyTorchВидео Using .backward() and Interpreting GradientsВидеоControlling Back Propagation of GradientsВидеоLab - Tensor Operations, Gradients, and GPU PracticeЛабораторная

Building Feedforward Neural Networks (FNNs)

Defining a Multi-Layer Perceptron with nn.Module and nn.SequentialВидеоRunning a Forward Pass and Exploring Model CapacityВидеоUnderstanding Loss Functions in Deep LearningЧтениеGetting Started with Optimizers: How Models LearnЧтениеBuilding the Training Loop for a Neural NetworkВидеоEvaluating Model Performance and Plotting ResultsВидеоLab - Train an MLP for Handwritten Digit ClassificationЛабораторная
03Optimization & Regularization for Deep Learning12 материалов

Regularization Techniques

Training Deep Models Isn't Just About More LayersВидеоWhat Is Overfitting & How Dropout and Weight Penalties HelpЧтениеL1/L2 in Practice and the Role of Batch NormalizationЧтение Applying Dropout to Prevent OverfittingВидеоUsing L2 Regularization with Weight DecayВидеоLab - Experiment with Regularization Techniques for Neural NetworksЛабораторнаяKnowledge Check - Regularization TechniquesЗадание

Weight Initialization & Optimization

Applying Custom Weight Initialization in PyTorchВидеоChoosing and Switching Optimizers in PyTorchВидеоWhy Initialization and Optimizer Choice MatterЧтениеLab - Experiment with Initialization and Optimizer CombinationsЛабораторнаяKnowledge Check - Initialization and OptimizationЗадание
04Convolutional Neural Networks (CNNs)16 материалов

Understanding CNNs

Why Convolutional Neural Networks Work So Well for ImagesВидеоUnderstanding Convolutions and Feature MapsЧтениеPooling, Activation & CNN vs. FNNЧтениеConvolution and Feature Maps — The Building Blocks of CNNsВидеоPooling, Padding, and ReLU — Understanding CNN TransformationsВидеоLab - Simulate a Convolution Operation with NumPy and Visualize FiltersЛабораторная

Building a CNN in PyTorch

Defining the Convolutional Layers of a CNNВидеоAdding Fully Connected Layers and Model SummaryВидеоPreparing and Training CNNs with PyTorchЧтениеTraining a CNN on MNISTВидеоEvaluating the CNN and Visualizing PredictionsВидеоLab - Implement and Train a CNN on CIFAR-10Лабораторная

Transfer Learning & Pre-trained Models

Loading and Customizing a Pre-Trained CNN for Transfer LearningВидеоTraining and Evaluating a Fine-Tuned CNNВидеоHow Transfer Learning Works and When to Use ItЧтениеLab - Fine-Tune a Pre-Trained Model on a New DatasetЛабораторная
05Recurrent Neural Networks (RNNs) & LSTMs14 материалов

Understanding RNNs

Why Deep Learning is Powerful for Sequential DataВидеоUnderstanding RNN ArchitectureЧтениеBPTT and Training Challenges in RNNsЧтение How RNNs Process Sequential Data: Concepts and Input FlowВидеоCharacter-Level RNN and Hidden State EvolutionВидеоLab - Build a Basic RNN to Model Sequential PatternsЛабораторная

LSTMs & GRUs

Getting Started with LSTMs in PyTorchВидеоHow LSTMs and GRUs Work InternallyЧтениеRunning Sequences and Comparing LSTM vs. GRUВидеоLab - Use an LSTM for Time Series Forecasting or Sequence ClassificationЛабораторная

NLP with RNNs & LSTMs

Text Generation with LSTMs in PyTorchВидеоSentiment Analysis with Hugging Face TransformersВидеоNLP Modeling: From Embeddings to TransformersЧтениеLab - Compare an LSTM Text Classifier with a Pre-trained TransformerЛабораторная
06Introduction to Generative AI19 материалов

Introduction to Generative AI

Welcome to Generative AIВидеоTraining a Discriminative Model: Logistic Regression on 2D BlobsВидеоFoundations of Generative AIЧтениеTypes and Use Cases of Generative AIЧтениеFitting and Visualizing a Generative ModelВидеоSample from a Simple Generative ModelЛабораторная

Evolution of Generative Models

Foundations of Generative Modeling: From GANs to VAEsЧтение From GANs to Autoregressive Models: Hands-On with Generative BasicsВидеоFrom Autoregressive to Diffusion: How Modern Generative Models Took OverЧтениеDiffusion Models in Action: From Noise to Realistic OutputsВидеоSample from a VAE and an Autoregressive ModelЛабораторная

Exploring LLMs & VLMs

Understanding Large Language Models (LLMs): Capabilities, Providers, and TrendsЧтениеWhat Can LLMs Do Today? Real Use Cases Across ProvidersВидеоUnderstanding Vision-Language Models (VLMs): Capabilities, Use Cases, and TrendsЧтениеWhat Can Vision-Language Models Do? Image + Text in ActionВидео

Ethical Considerations

Uncovering Bias in LLM OutputsВидеоHallucinations & Misinformation in ActionВидеоResponsible AI: Risks and Mitigation StrategiesЧтениеKnowledge Check - Ethical AI DeploymentЗадание
07Large Language Models (LLMs) & Transformer Architecture11 материалов

Understanding Transformer Architecture

Transformers Made LLMs Possible: Here's Why That MattersВидеоFrom RNNs to Transformers: A New Way to Process SequencesЧтениеAnatomy of Transformers and Their ArchitecturesЧтениеThe Problem with RNNs and How Transformers Fix ItВидеоSelf-Attention, Multi-Head Attention, and Feedforward NetworksВидеоTuning LLM Output with Temperature, Top-k, and Top-pВидеоExperiment with LLM Sampling ParametersЛабораторная

LLM Training & Fine-Tuning

LLM Fine-Tuning Strategies: From Supervised to AlignedЧтениеFine-Tuning a Transformer with Hugging FaceВидеоUnderstanding PEFT and Reinforcement Learning Fine-TuningЧтениеPerform Lightweight Fine-Tuning with LoRAЛабораторная
08Assessment2 материалов

Lesson

Learner Expectations for Skill AssessmentЧтениеSkill AssessmentЗадание