I,Planning Objectives

The Executive Yuan proposed the “Taiwan AI Action Plan” in January (107). In the future, the company will fully develop the AI application field with niche markets in China and will be able to shape Taiwan into an important hub for global wisdom and technology innovation. The Industry Bureau is guided by the 5+2 industry and service industry innovation needs and aims to cultivate talents in AI application skills and accelerate the development of innovative application services and industrial AI transformation.

This program provides AI talent training and subsidizes the cost of individual AI courses in a “subsidy” manner. If you are interested, you must first register your personal account with the program website. You agree to log in to the complete, detailed and authentic personal information at the time of registration, and you can apply for talent training after reviewing the application procedures. Scholars have access to teaching resources, and the program website also provides online and international online learning resources links (such as edx, course, Udacity, open university curriculum resources, MOOCs, etc.) to instantly grasp the pulsation of AI technology.

II, Talent Training Instructions

All AI training institutions are required to submit their own self-run courses to participate in the cooperation according to the training scope specified in this plan. After the course review, the course participants who meet the qualifications for applying for subsidy can refer to the individual website for the training needs. Courses, after enrolling in the course and completing the training, apply for the course training fee subsidy.
First, the subsidy object:

(1) A natural person who is 20 years of age and has a nationality of the Republic of China. (Under 20 years of age, in accordance with Article 12 of the Civil Code of the Republic of China, Article 13 (2) requires parental consent)

(2) In order to ensure the quality and effectiveness of the study, participants will be required to have basic or advanced skills in the AI course category:

1. Basic ability:

 
• Basic programming language (eg Python, C/C++, etc.)

 

• Industry know-how

2. Advanced ability:

 
• Introduction to calculus, linear algebra, probability statistics, etc.

 

• Interpretation of algorithms, machine learning, etc.

3. This program has 2 points or more of the learning points for the Roadmap related courses on the Taiwan AI educational platform.。
Reference website: https://idea.cs.nthu.edu.tw/~AIcoursemap/home/roadmap-interactive.html

(三) During the training period, those who receive government education and training subsidies may not apply for other government subsidies of the same type at the same time.

Second, the subsidy standard:

Subsidy recipients who meet the requirements of the first point will participate in the training course. The maximum annual subsidy of NT$ (the same below) is 30,000 yuan , :

(1) The training unit will handle the course fee standard and subsidize the training fee of up to 50% of each course. The rest of the training course participants will bear the burden.

(2) The training course is a special course that is in line with the policy. After the course is approved, it will be subsidized by 80% or full training.

(3) The proportion and amount of subsidies for each training course shall be announced after the approval of this Council according to the resolution of the course review meeting.

Third, the training unit qualification:

The qualifications of the training unit for the AI ​​Smart Application Training Course are as follows:

(1) Registered companies, consortium corporations; public and private colleges and universities; companies established under the company law, legally taxed and no refund records.

(2) Other qualifications shall be handled in accordance with the application for the plans of this Council.

Fourth, the training course review:

After accepting the plan submitted by the training unit, the board of directors should invite experts and scholars to form a review team for review, and after the approval, announce it to the project website for the recipients to apply for training. .

V. Course registration:

The registration method of this program is to adopt the online registration mechanism, and select the subsidy object to participate in the training qualification method. After registering the information on the online registration website, the online audit method is adopted. .

6. Subsidy review:

(1) The training subsidy object needs to sign the subsidy cut-off book, obtain the training certificate and the number of absenteeism does not exceed one-fifth of the total training time. Training grants.

(2) The training unit should, after 10 working days after the end of the course, check the following documents and send them to the Office for review, and then transfer the subsidy to the individual account or transfer to the training. Unit dialing:

1. Application list for grants and grants.

2. A list of attendance records and a check-in (return) form for the recipient.

3. A certificate of attendance or a certificate of credit.

4. The recipient’s payment receipt or original invoice.

5. The original application for the subsidy application.

(3) The services, such as various review or subsidies, as set out in this mechanism, may be entrusted to relevant authorities, legal persons or organizations.

Seven, training content description

(1) Training category

Class category	Course Hours	Training Program
Short term	Class hours total 12 hours or more   The course content needs to include the lecturer application technology demonstration or the subsidy object on the machine implementation	There are three categories of practical courses for AI application technology or advanced development:   AI Wisdom Application Class   AI core technology class   AI infrastructure class   For details, please refer to “II. Course Categories”
Themes Intensive course	Class with a total course time of at least 48 hours   There should be a specific vertical industry application theme and subsidy object implementation, the implementation time is at least 1/5 of the total number of courses	For the government to promote 5+2 key industries, such as smart healthcare, smart manufacturing and other smart application needs, plan thematic AI application intensive courses in specific vertical industries, including case materials and on-the-spot implementation.

(2) Course Area

Category	Training Program	Training purpose
AI Wisdom Application Class	AISmart Industry Apps Smart Manufacturing, Smart Healthcare, Smart Retail, Smart Finance, Smart Health Care, Smart Agriculture, etc.   AISmart Services app Product line monitoring, flaw detection, smart voice assistant, precision marketing, preventive maintenance, security monitoring, etc.   AISoftware and hardware integration application Industrial robots, human-computer interaction, self-driving, drones, smart wear, integrated applications, etc.	Enable application capabilities such as AI smart industry, AI smart service, and AI software and hardware integration
AI core technology class	AIOperations Machine Learning, Deep Learning, Reinforcement Learning, Generating Against Network (GAN), etc.   Computer Vision Image Processing and Identification, Object Detection and Recognition, etc.   Natural language processing Language comprehension, semantic analysis, cognitive services, voice conversations, etc.   Wisdom Reasoning and Data Exploration Recommended systems, data mining, data visualization, big data analysis, etc.	Enable technical skills such as AI algorithms, computer vision and natural language processing, and intelligent reasoning and data exploration
AI infrastructure class	AIPlatform Architecture Development Open source framework (Tensorflow, Caffe2, CNTK, MxNet) development and application, etc.   Big Data System Development Data acquisition, data preprocessing, data storage, distributed systems, etc.   AIChip and Sensor Development GPU parallel computing, embedded AI technology, edge computing, terminal device sensing technology integration, etc.	Enable development capabilities such as AI platform architecture, big data systems, AI chips and smart sensing systems