EVALUASI DECISION TREES DENGAN ALGORITMA C.45 BERBASIS OPTIMIZE WEIGHTS PARTICLE SWARM OPTIMIZATION UNTUK PENENTUAN LOKASI WARALABA
Jurusan Teknik Informatika, STT Nurul Jadid Probolinggo
Email : email@example.com
Franchises provide benefits to doing business under the banner of any other business that already has a good reputation. The idea, naming and management of a business have been in previous trials and are ready to implement the new location.
The strategic location is the placement of a company that can provide the maximum benefit to the company, because the purpose of the location strategy is to maximize profits for the company location. The failure caused the franchisor (the franchisor) beginner tempted by the great results without considering the factors that influence the location about business success .
It is necessary for consideration several factors before the franchise agreement. One of them is the prospect of choosing a business location , then use the model – based C4.5 decision tree algorithm Particle Swarm Optimization Optimize Weights (OW – PSO) This model will be used to create a rule and then used to predict the location chosen later.
It also made the selection criteria of information gain, gain ratio and Gini index in order to obtain a better rule. Experiments conducted with Rapid Miner yield 92.45 % accuracy for the criteria Gain Ratio, Information Gain criteria 88.68 % , and 90.57 % for the Ginny Index.
Key Word:: Waralaba, Algoritma C4.5, Optimize Weights Particle Swarm Optimization (OW-PSO), information gain, gain ratio dan gini index