Homeless prediction - Vestide

This extension analyzes your current vestide applications.

• Percentile ranking = (position/total * 100)

• IQR outlier detection: Tukey's method (Q1 - 1.5*IQR, Q3 + 1.5*IQR) [1]

• Z-score outlier detection: Values with |Z| > 2 are potential outliers [2]

• Pick top N things based on average percentile performance

• For pending application, use median number of total applications from processed list, and convert z-score into percentage probability where the true rank percentile is.

• Example:

• If pos is 5 out of 20 (top 25%), p_current = 0.25, n = 20

• Target threshold then is 20/200 = 10%

• Calculate how many std the current 25% is away from the target 10%

• In this case it's low

Version

Initial Release (v1.0)

• Popup dashboard to analyze housing application lists.
• Automatically separates 'Pending' and 'Processed' applications.
• Uses Z-score probability to predict the chance of finishing in the top 20 based on current position and historical applicant counts.
• Color-coded status (Green/Orange/Red) indicating the likelihood of success.

Guide

• Visit https://rooms.vestide.nl/en/my-vestide/my-applications/ and click the extension

Or:

• To use, visit room.vestide.nl
• Click "My vestide"
• Click "My applications" on the left side bar
• Click the extension

Test with python

• There's a python file to test separately. Go to applied houses site, copy everything and paste to asdf.txt. Then run main.py

Reference

[1] Tukey, J.W. (1977). Exploratory Data Analysis. Addison-Wesley

[2] Shiffler, R.E. (1988). Maximum Z scores and outliers; The American Statistician, 42(1), 79-80