Team Problem Solving (AI)
================

.. raw:: html
    
    <div style="background: #C3F8FF" class="admonition note custom">
        <p style="background: light-blue" class="admonition-title">
            Team Quiz: AI
        </p>
        <div class="line-block">
            <div class="line">This is a team problem solving session where each team solves a quiz.</div>
        </div>
        <ul>
            <li>Let's use what we've learned so far to find the answer to the quiz.</li>
            <li>It's time to find the answer to some of the quizzes below.</li>
            <li>After answering the team quiz, there will be presentations by each team.</li>
        </ul>
    </div>

.. raw:: html

Topic: 
    1. AI essential questions
        1. How AI systems handle complex decisions under uncertain circumstances, and how can techniques such as probabilistic modeling be applied to the autonomous driving field of robots?
        2. What is meta-learning, where AI systems learn how to learn, and how can this approach lead to the development of more adaptive and faster learning algorithms?
        3. How can AI use supervised learning to predict and prevent potential errors, and what role can it play in forecasting, maintenance, early disease diagnosis and risk mitigation?
        4. What is AI's ability to discover hidden patterns and structures in data through unsupervised learning and discuss applications in the areas of market analysis, anomaly detection, etc.?
        5. What are the roles of layers, neurons, in the way neural networks simulate the interconnected structures of the brain, and what are the similarities between how humans and machines learn?
        6. What is the mechanism of a convolutional neural network (CNN) and how it processes visual information layer by layer to recognize objects, faces and scenes?
        7. How can natural conversations and emotionally intelligent interactions be created in the mechanism of AI-based chatbots?

    2. AI choice questions
        1. Is there a way to apply DetectNet to situations beyond human detection, such as counting people in crowded areas or tracking animals for wildlife research?
        2. The importance of SegNet in semantic segmentation, where AI assigns a class label to each pixel in an image, and examples of how it can be applied to things like medical imaging, urban planning, and scene recognition?
        3. Is there a way to extend PoseNet's technology to estimate not only pose, but also depth information, contributing to 3D scene reconstruction and mixed reality experiences?
        4. In the problem of accurately segmenting a foreground subject in a complex background, how exactly does BackgroundNet's architecture solve this problem?
        5. What is DepthNet's role in estimating depth in 2D images so that artificial intelligence can detect the distance of an object from a camera, and how it can be applied to places like robotics, autonomous vehicles and 3D modeling?

    3. Nvidia Video


Evaluation Standard:
    1. Choose one AI essential question and one AI choice question to submit your report.

    2. Write a report about what you felt after watching the video.

    3. The report evaluation criteria are as follows.
        - A+ : Write at least 20 pages
        - A  : Write at least 15 pages
        - B+ : Write at least 10 pages