No abstract

No abstract

No abstract

Multiple Traveling Salesman Problem (MTSP) is a generalization of the Traveling Salesman Problem (TSP). MTSP is an optimization problem to find the minimum total distance of m salesmen tours to visit several cities in which each city is only visited exactly by one salesman, starting from origin city called depot and return to depot after the tour is completed. In this paper, K-Means and Crossover Ant Colony Optimization (ACO) are used to solve MTSP. The implementation is observed on three datasets from TSPLIB with 2, 3, 4, and 8 salesmen. Analysis of results using K-Means and Crossover ACO will be compared. The effect of selecting a city as depot on the total travel distance of tour will also be analyzed.

Stock portfolio is a kind of investment which consists of several stocks. The aim of a stock portfolio is to minimize the risk of an investment and maximize the return on investment. To construct the optimum portfolio of stocks, one needs a strategy of stock selection and must determine the percentage of investment in each stock selected. In this paper, both the priority index method and genetic algorithm are applied to optimize the stock portfolios in terms of the return. Priority index is used in stock selection based on some parameters: price/earnings (P/E), earnings/share (EPS), wealth creation, undervaluation, and price per earnings/growth (PEG). Stock selection in each sector is determined by choosing the stocks which have a priority index score at least equal to the minimum priority index score of the selected stocks. The minimum priority index score of the selected stock is determined by using a certain scale parameter. The percentage of investment in each selected stock is then determined by using a genetic algorithm. The results showed that increasing the value of scale parameters does not always increase the average return. Moreover, the stock selection with a wealth creation parameter has a higher average return than without a wealth creation parameter. Stock selection using daily data has a higher average return than annual data. The results also showed that the method has an optimum period of up to five months to make an investment decision.