Module 11

Question 1

nontraditional data sources

Answer 1

individuals (social media, online reviews etc)

busines process or corporate exhaust (retail scanner data)

IOT (smart building / vehicles)

Question 2

AI vs ML

Answer 2

AI = computer programs that possess decision making ability kinda like humans like neural networks

ML = extracting knowledge from data - finding and applying patterns - like an algo that identifies customers that are at risk of not renewing contracts

Question 3

overfitting

Answer 3

model too complex and treats NOISE as true param. low training data error, high test error

we focused too much on the training and validation!!! not enough testing

Question 4

underfitting

Answer 4

model too simple, treats true parameters as noise. High training data error, high test error

Question 5

Text analytics

Answer 5

analyze unstrucutred voice or text

Question 6

NLP

Answer 6

discern sentiment from nuance fo commentary