地點：Your Space 社群空間（近國父紀念館）

時間： 12/26 (四) 19:00 - 21:40 PM (6:40開放報到）

報名後，活動現場付現金 $100，千元大鈔請先換成百元鈔再過來喔。

若是學生，請報名後於現場報到出示學生證，可免費入場。

現場會備有輕食飲料。

特別感謝 數位時代 宣傳 及 Your Space 社群空間 場地贊助，謝謝凱爾和Claire及其團隊的幫忙！

也謝謝我們的夥伴 國泰金控 數位數據暨科技發展中心 贊助餐飲！

Agenda：

06:30-07:00 場地開放 & 報到

07:00-07:10 開場

07:10-07:40 The role of analytics in large scale ML model development, Ting-Po Lee 李庭柏, Senior Data Scientist@Netflix

07:40-08:10 Advocate Voices of Customers and Members through Data Science, Jason Wang 王致輝, Data Scientist@LinkedIn

08:10-08:20 中場休息

08:20-08:50 我們與資料科學的距離－國泰客服進線預測大解析, 陳冠穎, 資料科學分析師@國泰金控

08:50-09:20 現場Q&A

09:20-09:40 Networking

Introduction of Sharing:

👉 The role of analytics in large scale ML model development, Ting-Po Lee 李庭柏, Senior Data Scientist@Netflix

Abstract:

ML models are at the core of trillion dollar businesses today. From Facebook to Netflix, the success of their businesses stems from large scale ML models. Maintaining, evaluating and improving such models are challenging. In this talk, we will walk through some examples of how ML models actually work to generate real business value. Then talk about the challenges we face while developing them and how analytics can help in this ML model development process.

Bio:

Ting-Po Lee 李庭柏

Ting-Po is a senior data scientist in Netflix where he works on measuring and improving the performance of large scale personalization algorithms. Before Netflix he was a data scientist in Facebook ads ranking team where he used analytics to help the development of large scale machine learning models. Ting-Po received his MS in Statistics from Stanford University focusing on Data Science.

👉 Advocate Voices of Customers and Members through Data Science, Jason Wang 王致輝, Data Scientist@LinkedIn

Abstract:

LinkedIn is the #1 professional social network, currently with 660 million members, 30 million companies, 20 million open jobs, 90K schools, and 36K skills. In this sharing, Jason will give a case study of how we leverage data science to measure customer experience at scale and improve it through actionable insights. He will walk through the end to end process from data collection, preprocessing, model building (NLP, Text Mining), and driving business decisions. Finally, Jason will conclude the talk with the required skills when LinkedIn hires Data Scientists.

Bio:

Jason Wang 王致輝

Jason (Chih-Hui) Wang is a Data Scientist on the GCO (Global Customer Operations) Data Science Team at LinkedIn. He uses data to advocate the voices of customers and members. Prior to LinkedIn, he worked at LeanTaaS as a Data Scientist to help transform Healthcare Operations through Data Science. He holds a master’s degree in Statistics from the University of California, Berkeley and a bachelor's degree in Statistics from the University of National Chengchi University.

👉 我們與資料科學的距離－國泰客服進線預測大解析, 陳冠穎, 資料科學分析師@國泰金控

Abstract:

客服電話是企業跟消費者溝接觸的一個重要管道，身為全台最大、客戶數超過1,300萬的國泰金控來說，自然也擁有相當龐大的客服業務量，如何能更精準預測客戶進線的問題、減少客服的壓力，成為業務單位的當務之急！ 國泰金控數位數據暨科技發展中心的資料科學實驗室（Lab）在接到這個命題後，積極跟業務、商業分析團隊合作，從實務場景出發，分別以資料分析與工程的角度切入來優化客服進線預測，我們套用自然語言處理（Natural Language Processing，NLP）模型，並針對客戶的行為做顆粒度的重整，找出較泛化的行為顆粒度，更精準預測。 此次將會說明如何在模型的資料前處理、訓練資料集的切割方式，以及如何轉換成即時預測架構，讓大家身歷其境的了解，資料科學是場沒有終點的旅程，隨著資料科學專案的生命週期演進，我們將與資料科學越來越近～ 

Bio:

陳冠穎

現為國泰金控數位數據暨科技發展中心（數數發中心）資料科學實驗室的資料科學分析師，在大數據信評、客戶進線預測、國泰優惠贈禮關係等多個分析專案中，用圖學相關特徵工程、Spark 與 Neo4j 圖學運算、資料即時串流處理等方式來進行資料科學。有時作為偵探，地毯式的搜索可用資料源，並進行大量的資料探索；有時是資料夢想家，天馬行空地發想特徵與應用場景。