busn analytics- chapter 14

Decision Cycle

1. data

2. analysis

3. insight

4. decision

5. action

6. outcome

7. assessment

8. improvement

feedback loop

a control mechanism that is employed in economics, the sciences, and engineering to bring actual outcomes into alignment with desired outcomes

Process: input--> system--> output-->back to beginning

optimization

the overall goal of a cycle

negative feedback mechanism

mitigates fluctuations in the output by occurring in the direction opposite to the change.

perturbations

disturbances/ change in input

control system

the loop of constant feedback to adjust the system to achieve desired goals

positive feedback mechanism

the feedback loop adds to the input and occur in the same direction as the input

Challenges with data driven decision cycle

1. external events

2. unexpected events

3. competition

4. changing market forces

5. bad data

6. wrong decisions and actions

analysis-paralysis

- refers to the inaction that can result from overanalyzing without reaching a decision

- occurs when the analysis is too complicated, too intractable, and has too many variables

overfitting

when prediction models are fitted very tightly to the training data

- happens when noise and outliers are included in the training data instead of being ignored

attributes that analysts should have

- be open to criticisms and challenges to your assumptions

- be open to critiquing and challenging other analysts

- be open to recognizing unfavorable or unexpected insights in data

- be open to leaving no stone unturned (exploring the data from multiple viewpoints

- be open to admitting a lack of ability or knowledge

partly automated decision cycles

Utilize both human actions and computer automation

fully manual decision cycle

every stage of the cycle is assessed, evaluated, and moved forward by humans

Fully automated decision cycles

completed entirely by computers in all stages from data data acquisition to decisions to improvement

Artificial Intelligence (AI)

create systems that not only can be taught but also can self-learn by observing, experiencing, and operating in the real world

expert systems

rule-based and used to make decisions, particularly routine decision, with and without human help

-subset of AI

inference engine

interprets and evaluates facts to derive new rules based on its "experience" and expanding its knowledge

machine learning

utilizes AI, data mining, and statistics along with algorithms that can learn from data to make predictions

Examples of the Decision Cycle

- airline industry pricing

- Netflix recommendation engine

- baseball and the oakland A's

- ford motor and sustainable product design