CSCI 373: Predictions That Work

Michael A. Heroux

Any efforts in developing technology require regular assessment of where technical opportunities lie, what uses are important and which trends are changing. Knowing the fundamental drivers in a technology field (and nearby fields that may impact the value proposition of a given field) is very important. And making decisions that will lead to correct choices for new technology development is essential for success, especially for organizational leaders.

Challenge: As important as making predictions is, it is very hard to do, especially when you are new in a field. Also, good technology decisions are not based solely on technical issues; human behavior and social systems are major elements in making good predictions, as is being the right person in the right place at the right time with the right resources.

Approach: A universally valuable approach for accurate prediction is to develop models of the entities and relationships that are elements in your decision-making process. Models help you to identify and reason about the essentials and eliminate aspects that are secondary or irrelevant.

Here are a few resources for learning about prediction:

o Using Models to Make Predictions, National Geographic.

o What People Said About Books in 1498, John H. Lienhard.

o Making predictions with probability, Khan Academy.

o "State of Fear and Linear Prediction" Talk, by Michael Crichton.

Brief set of slides and exercises:

The Art of Prediction Michael A. Heroux

Assignment.

1. Read the National Geographic and Lienhard websites and summarize the content for class time discussion (spend no more than 30 minutes on this activity).

2. Go through the probability exercises on the Khan Academy (should take just a few minutes).

3. Watch the Michael Crichton talk above (52 minutes). Write a brief (few paragraphs, compact writing) reflection on this talk. Submit this reflection by email.