這份期末報告旨在 - 製作一份設計 AI(with ML&DL)的技術文件(2024年1月2日23:59 前交件)
評分標準(滿分最高 100)
25 pts - 獨創性
10 pts - 可行性
15 pts - 實用性
25 pts - Training plan 機器學習計畫(優化, 進化的計畫)
25 pts - Future expansion 未來擴增計畫
加分項目(歡迎挑戰):
1. Code for your product, 20 pts 的加分項目 Week 17(Jan 2nd 2024) 進行 demo
2. 2023 年 12/31 午夜前交件, 報告直接加分 5 pts.
Question for chatgpt (chatgpt 是個好夥伴)
How to construct a document to implement an AI with Machine learning and Deep learning technology?
Answer by chatgpt (參考 chatgpt 的回覆)
Constructing a document to implement an AI system with machine learning and deep learning technology involves several key steps. Below is a general guide to help you organize and structure your document. Keep in mind that the specifics may vary depending on your project, but this should provide a solid foundation.
1. Introduction: 引言
- Briefly describe the purpose and goals of the AI system.
- Provide an overview of the problem you are trying to solve.
- Mention the technologies you plan to use (e.g., machine learning, deep learning).
2. Problem Definition: 定義問題
- Clearly define the problem you are addressing.
- Discuss the significance of solving this problem.
- Identify any challenges or limitations.
3. Project Scope: 專案範圍
- Define the scope of your AI project.
- Specify any constraints(約束條件) or limitations(限制), such as data availability, computational resources, or time constraints.
4. Literature Review: 文獻探討
- Summarize existing research and solutions related to your problem.
- Highlight key methodologies(方法), algorithms(演算法), and models(模型) used by other researchers.
- Identify gaps(差距) in the current literature(文獻) that your project aims to address.
5. Data Collection: 資料收集
- Describe the data you will use for training and testing.
- Discuss the data sources, data types, and any preprocessing steps.
- Address issues related to data privacy and ethics(道德倫理).
6. Data Preprocessing: 資料的處理
- Detail the steps involved in preparing your data for training.
- Discuss techniques for handling missing data(缺失的資料), outliers(異常資料), and data normalization(正常化).
- Explain any feature engineering processes(特殊的資料處理工程技術...etc).
7. Model Selection: 模型選擇
- Justify(證明合法合適) the choice of machine learning and deep learning models.
- Explain how the selected models address the problem at hand. (明確且測底的解術你選擇的模型可以找出問題)
- Compare alternative models considered and explain why they were not chosen. (比較不同模型的不同並解釋你的選擇)
8. Model Training: 模型訓練
- Outline the training process, including hyperparameter tuning(包含參數的調整).
- Discuss any challenges encountered during training.
- Mention the evaluation metrics used to assess model performance(解釋你用來評估模型優劣效能的重點).
9. Model Evaluation: 模型評估
- Present the results of your model evaluation.
- Compare the performance of your model against baseline models or existing solutions(與目前現有或相近的模型做一簡短的比較).
- Discuss any limitations(限制) or potential biases(潛在的偏見[變數]) in your model.
10. Deployment: 發布成品
- Describe the process of deploying your model to a production environment.
- Discuss any scalability considerations(探討一下規模化的考量).
- Address issues related to real-time processing and inference(指出任何與即時運作與操作介面的可能問題).
11. Monitoring and Maintenance: 監控與維護
- Outline a plan for monitoring the deployed model's performance.
- Discuss strategies for handling model drift and updating the model over time(探討一下有關模型在運作一段時間後的調整策略).
12. Ethical Considerations: 道德考量
- Address ethical implications of your AI system.
- Discuss potential biases and steps taken to mitigate(緩解降低) them.
- Consider privacy concerns and measures implemented to protect user data(考量有關隱私問題的解決方案, 並探討如何實現並保護使用者的資料, big issue even openAI has the same issue).
13. Conclusion: 結論
- Summarize(總結) the key findings(重要的發現) and outcomes(成果結果) of your project.
- Discuss any lessons learned and opportunities for future work(探討在這個專案中所重新學到的觀點內容以及未來可能的其他發展機會).
14. References: 參考資料, 越多越好, 越有關越好
- Provide a list of all the sources and references cited in your document.
15. Appendix: 附錄
- Include any additional information, code snippets, or supplementary materials.
By following this structured approach, you can create a comprehensive document that guides the implementation of an AI system using machine learning and deep learning technologies. Adjust the details based on the specific requirements and nature of your project.