# CH1
Changing Business Environments
And Evolving Needs For Decision
Support And Analytics
• Big-bet, high-risk decisions.大賭注，高風險的決策。
• Cross-cutting decisions, which are repetitive but high risk跨領域決策，這些決策是重複性的，但風險很高需要小組合作。
that require group work.
• Ad hoc decisions that arise episodically.臨時出現的臨時決定
• Delegated decisions to individuals or small groups.將決定權下放給個人或小組

## 決策流程

• Define the problem 定義問題
• Construct a model 建立模型
• Identify and evaluate possible solutions 確定評估解決辦法
• Compare, choose, and recommend a solution to the
problem比較選擇推薦解決方法

A more detailed process is offered by Quain (2018):
1. Understand the decision you have to make.
2. Collect all the information.
3. Identify the alternatives.確定替代方案
4. Evaluate the pros and cons.評估利弊
5. Select the best alternative.
6. Make the decision.
7. Evaluate the impact of your decision.

## 內部、外部影響
Technology, I S, Internet, globalization, …技術，信息系統，互聯網，全球化
• Government regulations, compliance, …
– Political factors
– Economic factors
– Social and psychological factors
– Environment factors
• Need to make rapid decision, changing market conditions,

## Technologies for Data Analysis and Decision Support
• Group communication and collaboration
• Improved data management
• Managing giant data warehouses and Big Data
• Analytical support
• Overcoming cognitive limits 克服認知極限
• Knowledge management
• Anywhere, anytime support
• Innovation and artificial intelligence

# Simon’s Decision Making Process
– Includes three phases:
1. Intelligence
2. Design
3. Choice
4. [+] Implementation
5. [+] Monitoring

## Decision-making Processes
Phase 1 -
The Intelligence Phase: Problem (or Opportunity) Identification
• Issues in data collection 收集數據的問題
• Problem classification 問題分類
• Problem decomposition 問題分解
• Problem ownership 問題所有權
Phase 2 - The Design Phase設計
– Models
Phase 3 - The Choice Phase選擇
– Evaluating alternatives
Phase 4 - The Implementation Phase實施
– Implementing the solution
Phase 5 – Monitoring監控
• Phase 4 and 5 were not part of Simons’ original model

The Classical Decision Support ystem Framework 經典決策支持系統架構
• Degree of structuredness結構化程度
– Structured, unstructured, semistructured problems
• Type of control 控制類型
– Operational, managerial, strategic
• The decision Support matrix決策支持矩陣
• Computer support for …
– Structured decisions
– Unstructured decisions
– Semistructured problems

## Components of DSS 組件
• Data Management System
– DSS database
– Database management system (DBMS)數據庫管理系統
– Data directory 數據目錄
– Query facility 查詢工具
• The Model Management Subsystem 模型管理子系統
– Model base 模型庫
– MBMS
– Modeling language 建模語言
– Model directory 模型目錄
– Model execution, integration, and command processor模型執行，集成和命令處理器
• The User Interface Subsystem 用戶介面子系統
• The Knowledge-Based Subsystem 基於知識的子系統

## BI的架構
The architecture of BI
– Data warehousing (DW) [as a foundation of BI]
– Business performance management (BPM) 業務績效管理（BPM）
– User interface (dashboard)

## AI
• What Is artificial intelligence (AI)?
– Technology that can learn to do things better over
time.
– Technology that can understand human language.
– Technology that can answer questions.
• The major benefits of A I
– Reduction in the cost of performing work.
– Work can be performed much faster.
– Work is more consistent than human work.
– Increased productivity, profitability, …
## The Landscape of AI
• Major technologies
– Knowledge-based technologies 知識
– Biometric related technologies生物識別
• Tools and platforms …
• A I applications …
• Narrow (weak) versus general (strong) AI
• The three flavors of AI decisions
– Assisted intelligence輔助情報
– Autonomous AI自主AI
– Augmented Intelligence增強智能

## Societal Impacts of A I
• Impact on agriculture
• Contribution to health and medical care
• Other societal applications
– Transportation
– Utilities
– Education
– Social services

## data mining
是用人工智慧、機器學習、統計學和資料庫的交叉方法在相對較大型的資料集中發現模式的計算過程[1]。
## Data Mining Characteristics & Objectives
• Source of data for DM is often a consolidated 綜合data warehouse (not always!).
• DM environment is usually a client-server or a Web-based information systems architecture.
• Data is the most critical ingredient for DM which may include soft/unstructured軟/非結構化data.
• The miner is often an end user
• Striking it rich requires creative thinking
• Data mining tools’ capabilities and ease of use are
essential (Web, parallel processing, etc.)
### Types of patterns
– Association機會
– Prediction預測
– Cluster (segmentation)分群
– Sequential (or time series) relationships順序/時間序列關係
### 數據挖掘應用
• Customer Relationship Management客戶關係管理
–最大化營銷活動的回報
–提高客戶保留率（客戶流失分析）
–最大化客戶價值（交叉銷售，向上銷售）
–識別和對待最有價值的客戶
• Banking & Other Financial
–自動化貸款申請流程
–檢測欺詐交易
–最大化客戶價值（交叉銷售，向上銷售）
–通過預測優化現金儲備
• Retailing and Logistics零售物流
–優化不同位置的庫存水平
–改善店鋪佈局和促銷
–通過預測季節性影響來優化物流
–最大限度地減少由於有限的保質期造成的損失
• Manufacturing and Maintenance
• 經紀和證券交易
• 保險等等等無處不在

## Data Mining Process
– CRISP-DM (Cross-Industry Standard Process for Data資料探勘標準流程
Mining)
* Composed of six consecutive steps
– Step 1: Business Understanding了解業務
– Step 2: Data Understanding數據理解
– Step 3: Data Preparation數據準備
– Step 4: Model Building建立模型
– Step 5: Testing and Evaluation測試評估
– Step 6: Deployment部屬

– SEMMA (Sample, Explore, Modify, Model, and
Assess)代表“採樣”，“瀏覽”，“修改”，“建模”和“評估”。
– KDD 資料庫知識探索 (Knowledge Discovery in Databases)
提取是從結構化和非結構化來源創建知識。產生的知識需要採用機器可讀和機器可解釋的格式，並且必須以便於推理的方式表示知識。

# NGT
NGT程序通常在六個步驟中實施
階段：
與會者首先由會議主持人在
要討論的主題領域的初步陳述
然後，他們被引導對主題進行個別思考
小組主持人要求參與者陳述以下內容之一
他或她已經到達的回應
下一階段涉及整合和審核想法
然後要求他們確定相對重要性
應該與每個回應想法相符
最後階段是結果彙編
## NGT數據分析
NGT數據的分析涉及定性和定量分析
定量程序需要四個基本步驟：
將初始問題陳述分類為問題
主題
將概念模型中的問題主題重新組合為
形成主要問題維度
計算sc

## Delphi method
德爾菲法：調查組通過匿名方式對選定專家組進行多輪意見徵詢。調查組對每一輪的專家意見進行匯總整理，並將整理過的材料再寄給每位專家，供專家們分析判斷，專家在整理後材料的基礎上提出新的論證意見。如此多次反覆，意見逐步趨於一致，得到一個比較一致的並且可靠性較大的結論或方案。
## 腦力激盪法（英語：Brainstorming）
可以由一個人或一組人進行。參與者圍在一起，隨意將腦中和研討主題有關的見解提出來，然後再將大家的見解重新分類整理。在整個過程中，無論提出的意見和見解多麼可笑、荒謬，其他人都不得打斷和批評，從而產生很多的新觀點和問題解決方法。

## Herbert Alexander Simon
"bounded rationality"有限理性 and "satisficing".
preferential attachment
## Amos Tversky
human cognitive bias and handling of risk.處理風險、認知偏差
## Daniel Kahneman
cognitive basis for common human errors that arise from heuristics and biases啟發式和偏見引起的常見人為錯誤建立了認知基礎