MO Session 8

Limitations of predictive analytics (mistake coincidences)

2. Models can mistake coincidences for meaningful patterns ex. Super Bowl and the stock market

Big data effect on management

-Changes how we organize (ex: pokemon go, women's march) -Low cost platforms enable entrepreneurs to snap together an enterprise

Big data is changing the way we organize

-Firms no longer need brick-and-mortar operations; they just need a webpage and some algorithms to bring things together -Increasingly, firms, managers, and employees are becoming obsolete

Predictive Analytics

-The use of data, statistical algorithms, and machine learning techniques to identify the likelihood of future outcomes based on historical data >Benefits: Can model complex scenarios to improve decision-making, optimize processes, and help reach business goals

The 4 V's of big data

-Volume: massive quantities of info -Variety: diverse in type and structure -Velocity: rapidly generated -Veracity: data quality is not guaranteed, can be messy

Readings:

-they're watching you at work -in Unilever's radical hiring experiment -challenges and opps in managing human cap

Limitations of predictive analytics (overestimate accuracy)

1. Data analysts tend to overestimate the accuracy and predictive capabilities of their models

Limitations of predictive analytics (quality of inputs)

3. Models are only as good as the quality of the input data -Even if you have massive quantities of data, if measurement is poor or you are omitting important info, the model will be inaccurate

Limitations of predictive analytics (surprise events)

4. Even the best models cannot account for surprise events ex. 911

Limitations of predictive analytics (chain to past)

5. Models can be "overfit" to past data and mistake correlation with causation; they can chain us to our past

Webpage Enterprise

Coordinates inputs to create outputs through software and algorithms rather than via hierarchy, much as a webpage coordinates calls on resources to create the output on a browser EX: Circuit city

Big Data

Extremely large data sets that may be analyzed computationally to reveal patterns, trends, and associations, especially relating to human behavior and interactions