450 likes | 646 Views
Mixed Structural and Behavioral Models for Predicting the Future Behavior of some Aspects of the Macroeconomy. Mukund Moorthy 2nd February 1999. Contents. Economic Modeling System Dynamics Fuzzy Inductive Reasoning Proposed Macroeconomic Model Food Demand Modeling Conclusion.
E N D
Mixed Structural and Behavioral Models for Predicting the Future Behavior of some Aspects of the Macroeconomy Mukund Moorthy 2nd February 1999
Contents • Economic Modeling • System Dynamics • Fuzzy Inductive Reasoning • Proposed Macroeconomic Model • Food Demand Modeling • Conclusion
Economic Modeling • Economic Forecasting Techniques • Time Series Data • Neural Networks
Time Series Data • Time Series Components • Trend ( T ) • Cyclical ( C ) • Seasonal ( S ) • Irregular ( I )
Curve Fitting • Linear Trend Equation
Exponential Trend Equation Polynomial Trend Equation Curve Fitting
Smoothing Techniques • Moving Average • each point is average of N points • Exponential Smoothing
Time Series Forecasting • Box-Jenkins Method
Economic Forecasting • Step-wise Auto-regressive method • Neural Networks
System Dynamics • Modeling Dynamic Systems • Information feedback loops
System Dynamics • Levels • Flow Rates • Decision Functions
Levels and Rates Laundry List Levels Rates Inflows Outflows Population Birth Rate Death Rate Money Income Expenses Frustration Stress Affection Love Affection Frustration Tumor Cells Infection Treatment Inventory on Stock Shipments Sales Knowledge Learning Forgetting • Population • Material Standard of Living • Food Quality • Food Quantity • Education • Contraceptives • Religious Beliefs Birth Rate: System Dynamics
Forrester’s World Model • Population • Capital Investment • Unrecoverable Natural Resources • Fraction of Capital Invested in the Agricultural Sector • Pollution
Shortcomings of the World Model • Levels and Rates • Laundry List
Fuzzy Inductive Reasoning • Discretization of quantitative information (Fuzzy Recoding) • Reasoning about discrete categories (Qualitative Modeling) • Inferring consequences about categories (Qualitative Simulation) • Interpolation between neighboring categories using fuzzy logic (Fuzzy Regeneration)
Quantitative Subsystem Quantitative Subsystem FIR Model FIR Model Recode Recode Regenerate Regenerate Fuzzy Inductive Reasoning Mixed Quantitative/Qualitative Modeling
Modeling the Error • Making predictions is easy! • Knowing how good the predictions are: That is the real problem! • A modeling/simulation methodology that doesn’t assess its own error is worthless! • Modeling the error can only be done in a statistical sense … because otherwise, the error could be subtracted from the prediction leading to a prediction without the error.
Food Demand Model • Naïve Model • Enhanced Macroeconomic Model
Food Supply Food Demand Macroeconomy Population Dynamics Population Dynamics
Food Supply Food Demand Macroeconomy Population Dynamics Population Dynamics • Predicting Growth Functions k(n+1) = FIR [ k(n), P(n), k(n-1), P(n-1), … ]
Food Supply Food Demand Macroeconomy Population Dynamics Population Dynamics
Food Supply Food Demand Macroeconomy Population Dynamics Macroeconomy
Food Supply Food Demand Macroeconomy Population Dynamics Macroeconomy
Food Supply Food Demand Macroeconomy Population Dynamics Food Demand/Supply
Results • Annual / Quarterly Data • Layer One - Population Layer • Layer two - Economy Layer • Layer three - Food Demand Layer • Layer Four - Food Supply Layer • Optimization
Food Supply Food Demand Macroeconomy Population Dynamics Population Dynamics
Food Supply Food Demand Macroeconomy Population Dynamics Population Dynamics
Food Supply Food Demand Macroeconomy Population Dynamics Economy Layer
Food Supply Food Demand Macroeconomy Population Dynamics Food Supply Layer
Food Demand Layer Food Supply Food Demand Macroeconomy Population Dynamics
Conclusion and Future Work • Mixed SD/FIR offers the best of both worlds. • Application to any U.S. industry with change of demand and supply layers alone. • Application to any new country or region with new data for layers 1 and 2. • Fuzzy Inductive Reasoning features a model synthesis capability rather than a model learning approach. It is therefore quite fast in setting up the model.
Conclusion and Future Work • Fuzzy Inductive Reasoning is highly robust when used correctly. • Fuzzy Inductive Reasoning offers a self-assessment feature, which is easily the most important characteristic of the methodology. • Optimization with data collected at more frequent intervals.