Python has become one of the most popular programming languages: according to the latest TIOBE rankings, Python has surpassed C# and is now among the top 4 most popular languages globally alongside Java, C, and C++. The simplicity, readability, and extensibility of Python, along with its numerous extension libraries, make it an ideal development environment for engineers and researchers to handle experimental data, create charts, and even develop scientific computing applications. Currently, companies like Microsoft, Tencent, Google, Facebook, Baidu, and Alibaba are focusing on deep learning as a key area for future industrial and internet development. Research institutions and universities like the Chinese Academy of Sciences, Tsinghua University, and Peking University have established specialized research centers and laboratories to transform scientific and technological achievements in deep learning into practical applications, significantly promoting the development of deep learning applications.
To further promote research work in Python, artificial intelligence, machine learning, deep learning applications, and current practical projects in higher education institutions, research institutions, and enterprises, the Vocational Qualification Certification Training Center of the Chinese Management Science Research Institute, in collaboration with the Zhongke Software Research (Beijing) Science and Technology Center (http://www.fzby.org.cn/), invites frontline experts in the field of artificial intelligence academia and R&D to jointly hold a national training course on AI Python machine learning and deep learning. The course emphasizes a combination of theory and practice, focusing on hands-on operation; content is primarily code-driven, supported by theoretical explanations and formula derivations. The training is organized by the Vocational Qualification Certification Training Center of the Chinese Management Science Research Institute, with support from the Zhongke Software Research (Beijing) Science and Technology Center and Beijing Fuzhuo Baiyang Technology Co., Ltd.
1. Access to electronic handouts and data in advance after registration and payment for pre-study;
2. Participate in one training session, and you can attend the same online and offline courses for free for life;
3. After training, instructors provide students with their phone numbers and emails for post-training Q&A, ensuring effective outcomes;
4. Online classes will use the Tencent Meeting platform to ensure training quality, with sessions recorded for unlimited viewing;
5. Free trial available; please register at the end of the article for a trial arrangement;
6. Custom internal training is available, where instructors can come to your organization to train on specific topics, projects, and technical content of interest; please register at the end of the article for further coordination;
7. Limited to 40 people; early registration is encouraged; the first 20 registrants will receive recordings and materials from previous training sessions.
December 2, 2023 – December 4, 2023
On-site in Beijing / Tencent Meeting platform for 3 days
-
Using the Tencent Meeting platform for instruction, with years of online training experience to ensure quality. Recorded videos can be viewed unlimited times;
-
Access to electronic handouts and models in advance after registration and payment for pre-study;
-
After training, instructors provide students with their phone numbers and emails for technical support, ensuring effective outcomes post-training.
Senior experts from research institutions such as the Chinese Academy of Sciences and Tsinghua University. Frontline practical experts in AI, proficient in Python programming for artificial intelligence, focusing on various open-source projects in deep learning, such as TensorFlow, Caffe, and PyTorch. They favor a teaching style that combines theory and practice, with a clear and complete course structure that progresses from basic to advanced topics. They hold 2 patents and have completed numerous AI-related projects for schools, hospitals, enterprises, meteorological bureaus, and other institutions. They have been invited to provide internal training on AI technology for several large companies, including China Mobile, China Telecom, China Bank, Huaxia Bank, Pacific Insurance, State Grid, CNOOC, and Gree Electric, among others, including many Fortune 500 companies. The top IT training platform in the industry boasts a 99% satisfaction rate from 300,000 students;
|
|
Course |
Main Content |
|
Day 1 |
Chapter 1 Overview of Python Basics |
1、Setting up Python environment (Downloading Python, installation and version selection; Downloading PyCharm; Hello World in Python; installation and use of third-party modules; comparison of Python 2.x and Python 3.x) 2、Basic syntax of Python (Variable naming rules; basic mathematical operations; defining and manipulating common variable types; comments in Python) 3、Control flow in Python (Conditional statements; for loops; while loops; keywords break and continue; nested loops and variable loops) 4、Functions and objects in Python (Defining and calling functions; parameter passing and return values; variable scope and global variables; creating and using objects) 5、Installing Matplotlib and plotting (setting properties such as scatter plots, lines, axes, legends, annotations; plotting multiple graphs; nesting graphs) 6、Scientific computing libraries (Installing Numpy; creating ndarrays; indexing and slicing arrays; introduction to commonly used Numpy functions) |
|
Chapter 2 Introduction to PyTorch and Environment Setup |
1、Overview of deep learning frameworks (PyTorch, TensorFlow, Keras, etc.) 2、Introduction to PyTorch (Versions of PyTorch, dynamic vs. static computational graphs, advantages of PyTorch) 3、Installation and configuration of PyTorch (Pip vs. Conda package management, verifying successful installation) |
|
|
Chapter 3 Getting Started and Advanced Programming with PyTorch |
1、Definition of tensors (Tensors and their differences and relationships with scalars, vectors, and matrices) 2、Common properties and methods of tensors (dtype, device, layout, requires_grad, cuda, etc.) 3、Creating tensors (direct creation, creating from numpy, creating based on values, creating based on probability distributions) 4、Operations on tensors (addition, subtraction, matrix multiplication, element-wise multiplication, division, exponentiation, logarithm, approximation, clipping) 5、Indexing and slicing tensors 6、Understanding the autograd mechanism and computational graphs in PyTorch 7、Common toolkits and APIs in PyTorch (torchvision (transforms, datasets, model), torch.nn, torch.optim, torch.utils (Dataset, DataLoader)) |
|
|
Day 2 |
Chapter 4 Feedforward Neural Networks |
1、Basic principles of BP neural networks (What are the classifications of artificial neural networks? What is the difference between supervised and unsupervised learning? What is the topology and training process of BP neural networks? What is gradient descent? What is the essence of modeling BP neural networks?) 2、Implementation of BP neural networks in Python (How to split training and test sets? Why is normalization needed? Is normalization necessary? What are gradient explosion and gradient vanishing?) 3、Basic flow of implementing neural networks in PyTorch (Data, Model, Loss, Gradient) and training process (Forward, Backward, Update) 4、Case demonstration: Linear model, Logistic model, Softmax function output, BP neural network 5、Hands-on practice 6、Several issues worth studying (How to set the number of hidden layer neurons, learning rate, initial weights, and thresholds? What is cross-validation? Overfitting and underfitting, design of evaluation metrics, sample imbalance issues, model evaluation and selection (Occam’s Razor) etc.) |
|
Chapter 5 Convolutional Neural Networks |
1、Introduction to deep learning (Milestones in deep learning: Model + Big Data + GPU + AlphaGo) 2、Differences and relationships between deep learning and traditional machine learning (Is having more hidden layers in neural networks always better? What is the essential difference between deep learning and traditional machine learning?) 3、Basic principles of convolutional neural networks (What is a convolution kernel? What is the typical topology of CNNs? What is the weight sharing mechanism of CNNs? What features are extracted by CNNs?) 4、Building convolutional neural networks using PyTorch (Convolution layers, Batch Normalization layers, Pooling layers, Dropout layers, Flatten layers, etc.) 5、Case studies: Implementing object recognition using pre-trained CNN models 6、Hands-on practice 7、Introduction and quick start with PyTorch Lightning (Relationship between PyTorch Lightning and PyTorch, installation of PyTorch Lightning, case demonstration) |
|
|
Chapter 6 Transfer Learning |
1、Basic principles of transfer learning algorithms (Why is transfer learning needed? Why can transfer learning be done? What is the basic idea of transfer learning?) 2、Transfer learning algorithms based on deep neural network models 3、Case study: Cats vs. Dogs 4、Hands-on practice |
|
|
Chapter 7 Generative Adversarial Networks |
1、What are generative adversarial networks (GANs)? Why are GANs needed? What can GANs help us do? Insights brought by GANs) 2、Basic principles of GANs and the evolution of GANs 3、Case study: Implementing GANs in PyTorch (Generating handwritten digits) 4、Hands-on practice |
|
|
Day 3 |
Chapter 8 RNN and LSTM |
1、Basic working principles of Recurrent Neural Networks (RNNs) 2、Basic working principles of Long Short-Term Memory networks (LSTMs) 3、Case study: Time series prediction (Pollutant prediction in Beijing) 4、Hands-on practice |
|
Chapter 9 Temporal Convolutional Networks (TCN) |
1、Basic principles of Temporal Convolutional Networks (TCNs) 2、Differences and relationships between TCNs, 1D CNNs, and LSTMs 3、Case study: Time series prediction of COVID-19 4、Hands-on practice |
|
|
Chapter 10 Practical Case Study of Video Classification Based on Deep Learning |
1、Basic principles of video classification based on deep learning 2、Reading video stream files and extracting image frames 3、Using pre-trained CNN models to extract feature maps from specified layers 4、Custom construction of LSTM neural network models 5、Case study: Video classification with the HMDB51 dataset 6、Hands-on practice |
|
|
Chapter 11 Object Detection |
1、What is object detection? Differences and relationships between object detection and object recognition 2、Working principles of the YOLO model, differences between YOLO and traditional object detection algorithms 3、Case study: Implementing object detection using pre-trained YOLO models 4、Hands-on practice |
|
|
Chapter 12 Autoencoders |
1、What is an autoencoder? (Auto-Encoder, AE) 2、Introduction to several classic autoencoder models (AE, Denoising AE, Masked AE) 3、Case study: Feature extraction and reconstruction of handwritten digits using autoencoders 4、Hands-on practice |
|
|
Chapter 13 Review and Discussion |
1、How to search literature? (Do you know how to use Google Scholar, Sci-Hub, ResearchGate? Where should you look for data and code associated with papers?) 2、How to extract and explore innovative points? (If it is difficult to make original contributions at the algorithm level, how can you refine and explore innovative points based on your practical problems?) 3、Sharing and copying relevant learning materials (book recommendations, online course recommendations, etc.) 4、Establishing a WeChat group for later discussions and Q&A |
There are three types of training fees; please choose flexibly according to your needs.
Type A:3900 yuan/person (including training fee, material fee, Type A certificate fee, guidance fee, invoice fee, etc.)
Certificate: Advanced “Artificial Intelligence Application Engineer” professional capability certificate issued by Zhongke Software Research (Beijing) Science and Technology Center, which can serve as an important basis for the evaluation of professional technical personnel capabilities, salary increases, promotions, assessments, and appointments by relevant units.
Type B:4800 yuan/person (including training fee, material fee, Type A+B certificate fee, guidance fee, invoice fee, etc.)
Certificate: Advanced “Artificial Intelligence Application Engineer” professional capability certificate issued by the Vocational Education Research Institute of the Chinese Management Science Research Institute, which is included in the database of the Management Institute and can be checked nationwide, serving as valid proof for promotions and ratings.
Type C:5500 yuan/person (including training fee, material fee, Type A+B+C certificate fee, guidance fee, invoice fee, etc.)
Certificate: Advanced “Artificial Intelligence Application Management Engineer” professional skills certificate issued by the Ministry of Industry and Information Technology of the China Communications Industry Association (a national-level association), which can be checked nationwide and serves as valid proof for promotions and ratings.
Provides formal VAT invoices for easy reimbursement; if a meeting fee invoice is required, a meeting notice can be provided.
1. Students can enjoy a discount of 300 yuan with a student ID;
2. For groups of 3 or more (inclusive), each person can reduce 200 yuan;
3. For groups of 5 or more (inclusive), an additional free spot will be provided;
The above discount policies cannot be enjoyed simultaneously; only one can be applied.
Please scan the QR code below for online registration. After successful registration, we will send you a training notice and confirm by phone.
