致力於開發偽冒偵測AI模型,包含消費者異常登入行為、異常消費行為等等。主要在部門進行資料蒐集,資料視覺化,模型開發,系統開發及維運 工作內容 - 開發Flask API,提供其他單位串接以及蒐集資料 - 資料篩選, 資料清洗至模型建立, 並部屬成 docker 服務 - 開發資料即時同步 ElasticSearch 系統, 並且在 Kibana 上面呈現報表 - 使用Docker部屬所需之系統服務,例如: Redis, elasticsearch, kibana - 使用 Git 進行程式版本控制 - 協助團隊新技術導入 - 閱讀期刊會議論文,並且嘗試導入至可用模型服務 The department is committed to developing fraud detection AI models, such as anomaly login behavior, anomaly consumption behavior, etc. Mainly in the department for data collection, model development, system development, and maintenance work content - provide API to other departments to collect data - from data collection, and data cleaning to building AI models, and developing into docker services - Developed a real-time data synchronization ElasticSearch system, and presents reports on Kibana - Use docker to deploy system services required, such as Redis, ElasticSearch, Kibana - version control with Git - Assist the team in the introduction of new technologies - Read the paper, and do neural network research

- answer classmate questions - assist classwork questions - complete the tasks assigned by the professor - assist with final project examples and development

- 中央研究院計畫:以使用者有感為導向的環境物聯網大數據分析:以PM2.5為例 - Develop an estimation model to estimate time series Airbox data