BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20190719T085744Z
LOCATION:HG EO Nord
DTSTART;TZID=Europe/Stockholm:20190613T195000
DTEND;TZID=Europe/Stockholm:20190613T215000
UID:submissions.pasc-conference.org_PASC19_sess179_post139@linklings.com
SUMMARY:LIF02 - Using Iterative Random Forest to Develop Multi-Phenotypic 
 Genomic Selection Techniques
DESCRIPTION:Poster\n\n\nLIF02 - Using Iterative Random Forest to Develop M
 ulti-Phenotypic Genomic Selection Techniques\n\nRomero, Cliff, Kainer, Bro
 wn, Jacobson\n\nMachine learning algorithms, such as iterative Random Fore
 st (iRF), have potential in the area of Genomic Prediction both to minimiz
 e the number of required breeding generations to produce an ideal phenotyp
 e and as a way to optimize multiple phenotypes at once. IRF can be utilize
 d for its ability to distill combinatorial interactions from large sets of
  features, such as genetic markers, and associate them with a specific phe
 notype, such as drought resistance. After creating models for different ph
 enotypes, a set of potential parents can be selected that optimize all phe
 notypes concurrently. Using a <em>Populus trichocarpa </em>population, it 
 is possible to find the set of genetic markers that produces the optimal c
 ombination of heritable traits for biofuel/bioproduct production and to ac
 count for environmental impacts. By calculating random crosses between pot
 ential parents from the population, iRF models can be used to determine wh
 ich parents will produce a genotype closest to the predicted ideal set of 
 markers.  Since iRF models use the entire genome to make a prediction
 , and account for non-linear effects, it is possible to optimize multiple 
 phenotypes within a single population while minimizing the loss of predict
 ive power.
END:VEVENT
END:VCALENDAR