inflearn logo

Deep Learning with PyTorch Part 1: Mastering Basic Algorithms from A to Z

: "From Mathematical Foundations to the Latest Models, Completing the Deep Learning Pipeline with PyTorch (44 Lectures in Total)" The era of simply learning how to call model.fit() is over. From the very basics of artificial neural networks—differentiation and gradient descent—to essential PyTorch techniques used in the industry, and CNN/RNN for handling image and time-series data! We will help you systematically master the entire process of deep learning. Go beyond analyzed data and step into the amazing world of deep learning, where artificial intelligence mimicking the human brain learns and makes decisions on its own.

5 learners are taking this course

Level Basic

Course period Unlimited

Python
Python
AI
AI
Deep Learning(DL)
Deep Learning(DL)
PyTorch
PyTorch
Python
Python
AI
AI
Deep Learning(DL)
Deep Learning(DL)
PyTorch
PyTorch

What you will gain after the course

  • * A clear understanding of core mathematics for deep learning (calculus, chain rule)

  • * Architecture design skills from perceptrons to multi-layer perceptrons (MLP)

  • * Mastery of deep learning model implementation and deployment processes using PyTorch

  • * Professional model optimization skills using Dropout, Early Stopping techniques, and Optuna

  • * Basic proficiency in processing image and sequence data using CNN and RNN

Course Introduction

Deep learning is not a black box. Only when you understand how weights are updated internally and why backpropagation is necessary can you truly master and manipulate models at will.

This course will cultivate your ability to 'design' deep learning architectures. In particular, the progression from the principles of backpropagation in Lecture 23 to Optuna optimization in Lecture 36, and finally to transfer learning in Lecture 44, will elevate you from a simple developer to an AI expert. Upon completing this 44-lecture journey, you will gain the confidence to solve complex data problems, including images and text, using deep learning.


: "From Mathematical Foundations to the Latest Models, Completing the Deep Learning Pipeline with PyTorch (Total 44 Lectures)"

The era of simply learning how to call model.fit() is over. From the very foundations of artificial neural networks, such as differentiation and gradient descent, to essential PyTorch techniques used in the industry, and CNN/RNN for handling image and time-series data! We will help you systematically master the entire process of deep learning.

Now, move beyond analyzed data and enter the amazing world of deep learning, where artificial intelligence mimicking the human brain learns and makes decisions on its own.


Key points unique to this course

* Solid mathematical foundation: We select and easily explain only the essential math required to understand deep learning, from linear/quadratic functions to partial derivatives and the Chain Rule.

* Understanding Deep Learning Architecture: From Single-Layer Perceptrons to Multi-Layer Perceptrons (MLP), we thoroughly explore the principles of Backpropagation.

* The Pinnacle of Optimization Strategies: We will pass on everything from Dropout and Early Stopping techniques to prevent overfitting, to how to utilize Optuna, the latest hyperparameter tuning tool.

* Computer Vision & Sequence Data: Covers everything from MNIST image classification (CNN) and time-series/language data processing (RNN) to Transfer Learning for efficient training.

* Real-world Environment Setup: We have captured the entire practical process, from building a professional development environment using TensorFlow and Keras to model saving and automation.


Information about additional content can be found by visiting the Soft Campus website.

http://www.softcampus.co.kr/main.softcampus





📱 Curriculum & Project Preview


✒ Section 1. Deep Learning Overview and Mathematical Foundations (Lectures 1 to 9)

We will explore the history of deep learning and its fields of application, and cover the foundational mathematics (functions, derivatives, partial derivatives, and composite functions) that serve as the engine for neural network learning.

Key Learning: History of Deep Learning, Setting up Development Environment, Basics of Differentiation, Chain Rule


✒Section 2. Regression Models and Gradient Descent (Lectures 10 ~ 16)

We will study in depth the principles of linear regression, which is the basis of prediction, gradient descent for finding optimal weights, and the learning rate.

Core Technologies: Hypothesis Function Design, Error Evaluation Metrics, Multiple Linear Regression, Method of Least Squares

 

✒ Section 3. Principles of Classification Models and Perceptrons (Lectures 17 ~ 23)

Understand logistic regression and the sigmoid function, and learn the structure of the perceptron—the origin of artificial neural networks—and the multi-layer perceptron that solves the XOR problem.

Key Learning: Sigmoid function, Perceptron, Solving logic gates, Error backpropagation


✒ Section 4. Pytorch-based Neural Network Design (Lecture 24 ~ Lecture 30)

We will begin designing artificial neural networks in earnest using PyTorch. You will create your own optimal model structure by selecting activation functions and optimizers.

Key Technologies: How to use Pytorch, activation functions, optimizers, designing basic deep learning structures


✒ Section 5. Model Performance Optimization and Tuning (Lectures 31 ~ 36)

Learn the core practical techniques for preventing overfitting. You will study Dropout, Early Stopping, Model Checkpointing, and automated hyperparameter optimization using Optuna.

Key Learning: Checking for overfitting, Dropout, Early Stopping techniques, Model Checkpointing, and Optuna tuning



✒ Section 6. Advanced Architectures: CNN, RNN, and Transfer Learning (Lectures 37 ~ 44)

Master CNN, the core of image recognition, and RNN for handling sequence data through MNIST practice, and conquer transfer learning to reuse existing models.

Core Technologies: Binary/Multi-class Classification, CNN Architecture, MNIST Practice, RNN Basics, Transfer Learning



✒ Introduction to the Knowledge Sharer

Jaeseong Yoon (Lead Instructor of Data Analysis at Like Lion)


Development Experience
• Developed and launched SKT "Island Adventure" mobile content
• Developed and launched KT "Quiz Soccer" mobile content
• Launched SK "Mobile Real Estate Agent"
• Developed iPhone "Hanjatong" app
• Developed iPhone "Health Training" app
• Developed KT/SK Japanese Namco "Tales of Commons" content
• Developed KT mini-games (Yageum Yageum Land Grab, Aladdin's Magic Lamp, Mystery Block Detective Agency, BUZZ and BUZZ)

Teaching Experience
A veteran instructor with 19 years of experience in teaching and development for current employees of famous domestic companies and job seekers, including Samsung Multi Campus, Busan IT Industry Promotion Agency, Jeonju IT & Cultural Industry Promotion Agency, Incheon IT Industry Promotion Agency, Korea Radio Promotion Association, SK C&C, T Academy, Korea Institute for Robot Industry Advancement, Daejeon ETRI, Samsung Electronics, nica Education Center, Korea Productivity Center, Hanwha S&C, LG Electronics, and more.

Teaching Fields
Teaches in fields such as Java, Android, Frameworks, Databases, UML, iPhone, Big Data processing and analysis, Python, IoT, data analysis using R/Python, Deep Learning, Machine Learning AI, and Spark. The lectures are structured to explain concepts as easily as possible by incorporating diverse experiences, and to create examples that can be applied to practical exercises. Since this is not an offline class, please use the Q&A for anything you don't understand.

Recommended for
these people

Who is this course right for?

  • * Those who want to properly dig into the 'principles' of deep learning: Those who want to understand the mathematical background and backpropagation principles beyond just calling libraries.

  • * For those who want to handle PyTorch at a professional level: Those who want to master the entire process of designing and optimizing their own neural networks.

  • * Those eager to improve model performance: Those curious about practical techniques such as solving overfitting problems and hyperparameter tuning (Optuna).

  • * Those who want to get started with image and time-series data processing: Those who want to expand their AI portfolio by building a solid foundation in CNN and RNN.

Need to know before starting?

  • Basic knowledge of Python syntax, Numpy, and Pandas is required.

  • Even if you lack a mathematical background, the core concepts are covered within the lectures, so you can successfully complete the course as long as you have the passion.

Hello
This is softcampus

16,104

Learners

830

Reviews

595

Answers

4.7

Rating

46

Courses

Soft Campus is an education center that supports online and offline lectures and content sales.

For inquiries regarding AI-related fields and the purchase of various lectures and content, please contact us at raputa@nate.com or by phone at 02-553-0824.

Thank you.

More

Curriculum

All

46 lectures ∙ (17hr 13min)

Course Materials:

Lecture resources
Published: 
Last updated: 

Reviews

Not enough reviews.
Please write a valuable review that helps everyone!

softcampus's other courses

Check out other courses by the instructor!

Similar courses

Explore other courses in the same field!

Limited time deal

$22.00

48%

$42.90