Forums before death by AOL, social media and spammers... "We can't have nice things"
|    sci.environment    |    Discussions about the environment and ec    |    198,385 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 197,869 of 198,385    |
|    MrPostingRobot@kymhorsell.com to All    |
|    the waste export bidness (1/2)    |
|    15 Aug 21 19:39:44    |
   
   I have a little AI program that can answer 2 types of questions:   
   (a) which kinds of things result in X   
   and (b) what are the effects of X   
      
   It has a little built-in qualitative reasoning, some statistical   
   packages, and the ability to download databases off the net.   
      
   So I can ask it -- for those countries that important waste from the   
   developed world, what are the effects of those imports.   
      
   It found some quantitative data that was protected against upload (for   
   some reason some people keep tack of waste tonnage imported by various   
   countries, but then doesn't like to let anyone else find out what   
   those numbers are). But it also found a qualitative dataset.   
      
      
   Country Rank in waste imports bidness   
   bangladesh 9   
   china 5   
   ghana 1   
   guinea 13   
   guinea-bissau 17   
   haiti 14   
   india 6   
   indonesia 11   
   ivory_coast 10   
   kenya 12   
   lebanon 18   
   mexico 15   
   nigeria 3   
   pakistan 8   
   philippines 2   
   somalia 4   
   south 19   
   sweden 20   
   vietnam 7   
   zimbabwe 16   
      
      
   Fortunately, this is the kind of data its reasoning subroutines were   
   designed for. :)   
      
   Interesting, too, that Sweden appears No 20 on the list. Sweden itself   
   generates so little waste it has to import garbage from other   
   countries to burn in its garbage-fired electricity generator network.   
      
   So we can take this list and ask Mr AI -- what are the effects of   
   importing more waste?   
      
   This is its coded output:   
      
   Effect R2 Beta stderr(Beta)   
   jntsv3 0.75952576 0.596365 0.205042   
   pubsrv3 0.72034981 0.960609 0.365721   
   pubsv 0.71756747 0.782357 0.299915   
   services 0.65874095 1.60003 0.703688   
   popden 0.65616233 -10.9573 5.21477   
   yag 0.65443842 -0.0401269 0.0191701   
   pubcon 0.62395690 1.03402 0.490497   
   comin4 0.61757346 1.1873 0.570897   
   comin3 0.61309658 2.48082 1.2042   
   ppubco 0.61308916 1.24042 0.602116   
   pprico 0.61308890 -1.24042 0.602116   
   max-ind-tax 0.57101594 -0.933333 0.438042   
   yprisrv 0.39497001 0.0416132 0.033861   
   yhhinc 0.35940589 3.43178e-05 3.28081e-05   
   ywphh 0.34794705 -1.12386 1.01148   
   ypimpcp 0.34167942 0.0025586 0.00254316   
   netexp 0.33300236 0.36895 0.343296   
   yzlit 0.32498972 0.0155146 0.0147002   
   ygdpwnw 0.29207130 3.72298e-05 4.15054e-05   
   yzlifewom 0.28541485 0.0348371 0.0362404   
   yzpop 0.27980459 -0.000774238 0.000816647   
   ydth65 0.26315863 -0.00180884 0.00198994   
   gdpperca 0.26070686 0.082918 0.0800293   
   covidcasesfeb2021 0.25970882 6.29817 4.81305   
   yzlifemen 0.25861037 0.0410799 0.045729   
   coviddeathsfeb2021 0.25491355 0.124608 0.100283   
   prisrv3 0.24826431 0.639419 0.679874   
   prod 0.23190030 1.99594 2.38818   
   prisv 0.22989798 0.707249 0.790946   
   ypubdt3 0.20268407 0.030949 0.035182   
   yimpcap 0.18990743 0.000157035 0.000198181   
   ytrdcap 0.18976745 7.60843e-05 9.60631e-05   
   yexpcap 0.18959945 0.000147566 0.000186417   
   ypubdebt 0.17659647 0.041256 0.0482376   
   yjntdt3 0.14194186 0.0351416 0.0495214   
      
   The "R2" column tells you how much of the effect waste imports seem to   
   affect. So for the first line in the table (that is the% of GDP   
   that services shared between the public and private sector) about 75%   
   of the country-to-country variation in joint services are "predicted"   
   by similar country-to-country variation in waste imports (as indicated   
   by the rank -- where smaller numbers == higher levels of waste imports   
   and higher numbers == lower levels of waste imports).   
      
   The Beta column describes how much of the effect is affected by   
   changes in waste imports (rank). For joint services each increase in   
   rank (i.e. reduction in waste imports) results in about 0.6% more GDP   
   spent in public/private services.   
      
   The following lines describe other statistics to do with services in   
   the economies concerned. In all cases more waste imports result in   
   lower levels of services.   
      
   "popden" shows more waste predicts lower population density. More   
   waste imports seem to predict the relevant country is very crowded.   
      
   "netexp" shows those countries with more waste imports are the ones   
   that have the lowest net exports as pct GDP. Essentially   
   waste-importing countries are not making any money in the   
   international trade scene. The more they import, the less they   
   make. No wonder they have no govt services.   
      
   More waste imports also predicts lower literacy, lower female life   
   expectancy, lower male life expectancy, lower productivity, lower GDP   
   per capita, and higher death rates for people over 65.   
      
   Big waste importers have lower levels of international debt. Most   
   likely because no-one will lend them any money for development.   
      
   On the plus side waste importing countries have almost zero tourism   
   which means they have the lowest levels of COVID.   
      
   For the male lifex each rank in waste importing has 1/4 of a year   
   longer life. Since there are 20 ranks ("top20") this means the biggest   
   waste importers have 5 years less life for men than e.g. Sweden (at   
   rank 20). 5 years of lifex may not sound much but it's the difference   
   between being a "first world" country and a "pest hole".   
      
   The AI ensures all of these relationships are statistically   
   significant in at least 2 tests at 90% confidence. IOW there is about   
   1 chance in 10,000 these links found are just chance patterns in the   
   data. The links are "real".   
      
   Visit my kaggle page at
|
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca