Stephanie's Extended Abstract

Abstract

• Not sure on the phrase "In the scope of senior thesis"

Predictive Analytics

• Overview
  • change "ever since 1940s." to "ever since 1940." or ever since the 1940's.
  • change "During 1950s and 1960s" to "During the 1950's and the 1960's."
  • And is used twice while listing things in this sentence: "During 1950s and 1960s, predictive analytics got involved in more fields, mostly in big corporations and research institutions, such as weather (ENIAC weather forecasts), air travel and logistics (shortest path problem solved), and banking (FICO using predictive modeling for credit risk decisions)."
• Process
  • I would use the numbers when referring to a section not just the name
• Applications
  • The words in figure 1 are a little hard to read
  • I would start the first sentence with "The"
  • Explain figure 1 more, what are the percents percents of?

Random Forests

• Overview
  • not sure if the "certain limit as the number..." is clear.
  • Would be helpful to have image of a random forest
• Implementation Process
  • Can't read words on the figure
  • Not clear on what bootstrapping is. Are you saying that it is both of the steps or just the averaging? I also believe you could go into more detail about bootstrapping like how "Each sample is different from the original data set, yet resembles it in distribution and variability. "[wiki]
  • I would quote the Gini impurity part since we have not researched and tested every possible way to determine a split
  • I would say at the beginning of describing the process of building a simple decision tree that you took a subset of the data for training. Also, I would go into the bootstrapping here since each tree is trained on a different subset of the data
• Applications
  • not sure what you mean when you say "compatible data sets"
  • I would change up the first sentence a little bit. I got confused reading it and had to read it multiple times

Project Methodology

• Data set
  • Table is hard to read
  • How/where were you able to get the data set?
  • The title Once Upon a Crime: Towards Crime Prediction from Demographics and Mobile Data should be either quoted, underlined, or italicized based on if it was an article, book, etc.
  • "on such data set" to "on such a data set"
• Related Work and Current Approach
  • What exactly is IUCR?
• Project Progress
  • How are you comparing the algorithms' pefermances?
  • How are you determining success of your own algorithm? Why only accuracy and not the other ways of determining success?
  • How do you define improve in terms of improving your algorithm?