本課程目的在透過人工智慧結合物聯網(AIOT)、腦科學、量子機器學習、磁性系統相位辨識等專題分組實作，訓練同學跨領域應用人工智慧解決物理問題的能力。
本課程內容包括:
一、AIOT(含自駕車)專題
1. 3D繪圖與列印
2. 樹莓派控制板與Linux作業系統
3. 感測器,GPIO與藍芽無線控制
4. Python程式設計
5. OpenCV電腦視覺程式應用
6. Tensorflow、Keras 機器學習人工智慧應用
7. 人工智慧結合物聯網(AIOT)主題實作
二、腦科學專題
1. 大腦影像資料結構
2. 以卷積神經網路進行大腦影像處理：超大型影像拼接（stitching）、神經元切割（segmentation）與追跡（tracing）
3. 以非監督式學習進行大腦連結體網路分析
4. 上述腦科學演算法之整合使用者介面
三、量子機器學習專題
1.量子計算基礎-量子線路模型
2.將古典數據轉化成量子態中的數據
3.量子演算法-量子傅立葉變換、量子搜尋
4.量子神經網絡，量子主成分分析算法，量子支持向量機
5.實作：量子圖像辨識 ，利用量子電腦模擬分子的基態能量
四、磁性系統相位辨識
1.磁性模型
2.蒙地卡羅模擬
3.深度學習方法: 卷積神經網路(CNN)、自编碼器(autoencoder)
4.使用深度學習進行相位辨識
The purpose of this course is to train students to apply artificial intelligence to solve physical problems across domains through the sub-assembly of topics such as artificial intelligence network (AIOT), brain science, quantum machine learning, and magnetic system phase recognition.
The course includes:
1. AIOT (including self-driving cars) topic
1. 3D drawing and printing
2. The Spiral Pi Control Board and Linux Operating System
3. Sensor, GPIO and blue bud wireless control
4. Python Programming
5. OpenCV computer visual application
6. Tensorflow, Keras Machine Learning Artificial Intelligence Applications
7. Artificial Intelligence Compound Network (AIOT) Theme Implementation
2. Brain Science Specialties
1. Brain image data structure
2. Use the coil neural network to perform brain image processing: super-large image stitching, neural segmentation and tracing
3. Use non-supervised learning to conduct brain-connected network analysis
4. Integrated user interface of the above-mentioned brain science algorithm
3. Quantum Machine Learning Specialties
1. Quantum computing foundation-quantum circuit model
2. Transform classical data into data in quantum state
3. Quantum algorithm-Quantum Fuliye transformation, quantum search
4. Quantum neural network, quantum principal component analysis algorithm, quantum support vector machine
5. Implementation: Quantum image recognition, using the basic energy of quantum computers to simulate molecules
4. Magnetic system phase identification
1. Magnetic model
2. Monte Carlo Mockup
3. Deep learning method: Volume neural network (CNN), self-coders (autoencoder)
4. Use deep learning for phase identification
教學講義
https://physexp.thu.edu.tw/~AP/YC/AIcar/
Teaching Lecture
https://physexp.thu.edu.tw/~AP/YC/AIcar/
| 評分項目 Grading Method | 配分比例 Grading percentage | 說明 Description |
|---|---|---|
| 期末考期末考 Final exam |
100 |
