- This paper introduces the Normal Boundary Intersection (NBI) method for highly accurate Pareto frontier prediction in multicriteria optimization as compared to the traditional Weighted Sum Scalarization technique.
- It presents dynamic portfolio optimization, highlighting its advantages over static methods in finance.
- It states the mean-variance Portfolio model and introduces a new systematic procedure employing NBI for dynamic portfolio optimization through strategic asset reallocation.
- Additionally, it showcases empirical analysis using curve fitting techniques to predict future stock prices by defining a polynomial function that can predict the price of a stock given any time in the future.
There are 3 notebooks:
- NBI_optimization: this introduces the cocept of the NBI method and compares it with the traditional Weighted Sum Scalarization with the help of an optimizaation problem.
- Portfolio_Optimization_NBI: This is the code to the dynamic portfolio allocation strategy using NBI method reciprocated on 3 stocks of the NIFTY 50.
- StockPricePrediction_Regression: Uses curve-fitting techniques to predict stock prices. In the code, I have used a 3-degree polynomial and a 10-degree polynomial and found the optimal parameters for the best-fitting curve.