...

Jorge A García

OPTIMIZATION - ECONOMIC MODELING - DATA SCIENTIST

ABOUT ME

I'm an optimization modeler and data scientist with more than 10 years of experience developing modeling solutions and obtaining insights from large datasets (> 10 GB). My intersdiciplinary background includes a PhD in Systems Design Engineering, a MSc. in Operations Research and a BSc. in Industrial Management.

I have extensive hands-on experience developing a wide range of decision models and tools. Examples of these are the following:



OPTIMIZATION MODELS & ALGORITHMS


  • Spatial Mixed-Integer Linear Program for Nutrient Measure Selection
    Formulated and implemented a mixed-integer linear program that determines the least-cost way to achieve an environmental nutrient target by selecting optimal abatement measures for different sub-watersheds. The paper is here.

  • Multi-Objective Optimization for Water Resources Management
    Formulated and implemented a goal programming model that allocates water volumes from various water supply sources (groundwater, surface water, desalinated water, treated wastewater) to different water uses (domestic, agricultural, industrial), taking into account stakeholders' prioritization. The publication is found here and the model here.

  • Quadratic Programming for Optimal Investment in Abatement
    Formulated a quadratic program that builds up from a hydrological steady-state model of the Lake Erie watershed and determines the level of investment in wastewater treaatment plants or agricultural areas such that a target phosphorus reduction is achieved. Preprint found here.

  • Integer Programming for Production Planning
    Formulated and implemented an integer program for production planning on a manufacturing company. This model was developed as part of my MSc. thesis which and can be found here.

MACHINE LEARNING & DATA SCIENCE


  • Machine Learning Applied to Discrete Choice Surveys
    Applied machine learning to predict the willingness-to-pay of respondents about beach clean-up measures from a survey applied to several European countries. Results are comparable to those from traditional discrete choice models.

  • Land Use Clustering
    Grouped census sub-divisions in Ontario based on their land use shares and socio-economic variables using K-means clustering. This grouping will be used in a model to predict future land use changes.

  • Integrated Choice and Latent Variable
    Specified and solved an integrated choice and latent variable model about attribute-non-attendance on responders of a survey. The novelty of this approach is that I used PCA and ICA to select the socio-econ variables from which latent variables are created.

  • Canadian Census Data Processing
    Developed a procedure to extract census data for multiple years and load them into SQLite databases for posterior use in modeling and analyses. Code will be soon made available on GitHub.

MACROECONOMIC MODELING


  • Computable General Equilibrium
    Developed a CGE model for the Canadian economy to evaluate the impact of future climate-induced water shocks. The model has 49 industries, 64 commodities, and 4 economic agents. It was constructed in Python but the core solver is from GAMS. The preprint is here and the model details here.

  • Canadian Social Accounting Matrix
    Developed a procedure to automatize the creation of Social Accounting Matrices for the Canadian economy using tables from Statistics Canada. Tables are found here.

  • Multi-Regional Input Output models
    Developed a hydro-economic model combining the Input Output theory and linear programming. This model was used to evaluate the cost of water disruptions or emission constraints into the Great Lakes economy. Publications are here: WRR and ESR.

  • Discrete Choice Models
    Formulated and estimated discrete choice models related to environmental or water quality surveys. For example, I participated in the specification of a model to describe respondents' preferences and willingness-to-pay about beach litter cleaning measures.

SYSTEMS ENGINEERING


  • Reaction-Diffusion Equation for Containers' Lifetime Calculation
    Numerically solved a reaction-diffusion system to approximate the reduction of sulphate to sulphide caused by sulphate-reducing bacteria under the conditions of a deep geological repository in order to calculate the corrosion rate and therefore the expected lifetime of used nuclear fuel containers. The publication is found here.

  • Agent-Based Model of Drivers in a Road Network
    Formulated an agent-based model to describe the effect on road networks of drivers using real-time information about current travelling times to choose their destination route. The model is presented here.

  • Bayesian Belief Updating Procedure
    Developed a procedure to assimilate new information and update current beliefs about the probability distribution of the true state of a system. This procedure was applied to updating travel time estimation in a road network. The procedure can be found in Chapter 6 of my PhD thesis (link here).

  • Discrete-Event Simulation of a Production Process
    As part of my MSc. thesis, I formulated a discrete-event simulation model of a real production process. Model details are found here: here.

CONTACT

ja4garci@uwaterloo.ca
Kitchener, ON, Canada
GitHub: https://github.com/jorge-antares

Click the button to show where you are! (takes a few seconds)