BEGIN:VCALENDAR VERSION:2.0 X-WR-CALNAME:EventsCalendar PRODID:-//hacksw/handcal//NONSGML v1.0//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York LAST-MODIFIED:20240422T053451Z TZURL:https://www.tzurl.org/zoneinfo-outlook/America/New_York X-LIC-LOCATION:America/New_York BEGIN:DAYLIGHT TZNAME:EDT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZNAME:EST TZOFFSETFROM:-0400 TZOFFSETTO:-0500 DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT CATEGORIES:College of Engineering,Lectures and Seminars,Thesis/Dissertation s DESCRIPTION:College of Engineering Data Science MS Thesis Defense Title: Em ulating Marine Ecosystem Dynamics: A Machine Learning Approach for Biomass Prediction and Forecasting by Ajmal Abbas Date & Time: Monday, May 11, 20 26 at 4:00 PM Meeting Link: Join: https://teams.microsoft.com/meet/2101346 22120814?p=HF1gXRarGipViGBpda Meeting ID: 210 134 622 120 814 Passcode : Ut7pS6RH Committee chair: Dr. Gokhan Kul Committee Members: Dr. Gavin Fa y (co-advisor) Dr. Ashok Kumar Patel Dr. Firas Khatib Abstract: Marine Eco system models such as Atlantis provide valuable insights into multi-specif ic interactions, spatial dynamics and fisheries management, but their high computational cost limits rapid scenario analysis and real time decision- making. This study presented a data driven ecosystem emulator for the Nort heast U.S. Atlantis model using machine learning and deep learning techniq ues. A large-scale dataset (~3.4 million records) spanning 1964-2020 was c onstructed from over 500 Atlantis model run scenarios, integrating biomass of species functional groups (guilds), spatial polygon, fishery removals and environments variables including temperature, salinity and primary pro duction. The Emulator adopted a hybrid modeling framework combining Random Forest regression for accurate biomass prediction and Long Short -Term me mory Networks for temporal forecasting of ecosystem dynamics. Feature Engi neering incorporated spatial biomass distribution, lagged variables to cap ture ecological inertia and time aware transformations. Model evaluation i ncluded out of sample temporal validation, recursive back testing and ecol ogical plausibility assessments to ensure robustness and realism. Result d emonstrated strong predictive performance, with guild-level R2 values freq uently exceeding 0.90 and overall model accuracy approaching 94%. The fore casting component produced stable, biologically consistent trajectories wi thout unrealistic oscillations. Importantly, the emulator achieved high co mputational efficiency, with end-to-end prediction and forecasting complet ed in under few seconds, substantially reducing runtime compared to the or iginal Atlantis simulations. This work established a scalable and efficien t alternative to process-based ecosystem models, enabling rapid scenario t esting and supporting data -driven fisheries management and environmental decision making. Meeting Link: Join: https://teams.microsoft.com/meet/21 0134622120814?p=HF1gXRarGipViGBpdaMeeting ID: 210 134 622 120 814 Passco de: Ut7pS6RH\nEvent page: https://www.umassd.edu/events/cms/data-science-m s-thesis-defense-by-ajmal-abbas.php\nEvent link: https://teams.microsoft.c om/meet/210134622120814?p=HF1gXRarGipViGBpda X-ALT-DESC;FMTTYPE=text/html:

College of Engineering

Data Science MS Thesis Defense

Title: Emulating Marine Ecosystem D ynamics: A Machine Learning Approach for Biomass Prediction and Forecastin g

by Ajmal Abbas

Date & Time: Monday\, May 11\, 2026 at 4: 00 PM

Meeting Link: Join: https://teams.microsoft.com/meet/21 0134622120814?p=HF1gXRarGipViGBpda

Meeting ID: 210 134 622 120 814 Passcode: Ut7pS6RH

Committee chair:

\ n

Dr. Gokhan Kul

Committee Members:

Dr. Gavin Fay (co-advisor)

Dr. Ashok Kumar Patel

Dr. Firas Kh atib

Abstract: Marin e Ecosystem models such as Atlantis provide valuable insights into multi-s pecific interactions\, spatial dynamics and fisheries management\, but the ir high computational cost limits rapid scenario analysis and real time de cision-making. This study presented a data driven ecosystem emulator for t he Northeast U.S. Atlantis model using machine learning and deep learning techniques. A large-scale dataset (~3.4 million records) spanning 1964-202 0 was constructed from over 500 Atlantis model run scenarios\, integrating biomass of species functional groups (guilds)\, spatial polygon\, fishery removals and environments variables including temperature\, salinity and primary production. The Emulator adopted a hybrid modeling framework combi ning Random Forest regression for accurate biomass prediction and Long Sho rt -Term memory Networks for temporal forecasting of ecosystem dynamics. F eature Engineering incorporated spatial biomass distribution\, lagged vari ables to capture ecological inertia and time aware transformations. Model evaluation included out of sample temporal validation\, recursive back tes ting and ecological plausibility assessments to ensure robustness and real ism. Result demonstrated strong predictive performance\, with guild-level R2 values frequently exceeding 0.90 and overall model accuracy approaching 94%. The forecasting component produced stable\, biologically consistent trajectories without unrealistic oscillations. Importantly\, the emulator achieved high computational efficiency\, with end-to-end prediction and fo recasting completed in under few seconds\, substantially reducing runtime compared to the original Atlantis simulations. This work established a sca lable and efficient alternative to process-based ecosystem models\, enabli ng rapid scenario testing and supporting data -driven fisheries management and environmental decision making.

M eeting Link: Join: https://teams.microsoft.com/meet/21013462212081 4?p=HF1gXRarGipViGBpda
Meeting ID: 210 134 622 120 814
Pa sscode: Ut7pS6RH

Event page: https://www.umassd. edu/events/cms/data-science-ms-thesis-defense-by-ajmal-abbas.php
Ev ent link: https://teams.microsoft.com/meet/210134622120814?p=HF1gXRa rGipViGBpda

DTSTAMP:20260422T231741 DTSTART;TZID=America/New_York:20260511T160000 DTEND;TZID=America/New_York:20260511T170000 LOCATION:Teams SUMMARY;LANGUAGE=en-us:Data Science MS Thesis Defense by Ajmal Abbas UID:6155b0eb09f74e4f30930dc842476dae@www.umassd.edu END:VEVENT END:VCALENDAR