Optimal Prediction of Foreign Exchange Market Exploring the Effect of Water Cycle Algorithm Strategy on Data Segregation
Arup Kumar Mohanty1, Debahuti Mishra2 

1Arup Kumar Mohanty, Department of Computer Science and Information Technology, Siksha O Anusndhan University, Bhubsneswar, Odisha, India.
2Debahuti Mishra, Department of Computer Science and Engineering, Siksha O Anusndhan University, Bhubsneswar, Odisha, India.

Manuscript received on 06 March 2019 | Revised Manuscript received on 13 March 2019 | Manuscript published on 30 July 2019 | PP: 4738-4748 | Volume-8 Issue-2, July 2019 | Retrieval Number: B3753078219/19©BEIESP | DOI: 10.35940/ijrte.B3753.078219
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: The market of foreign exchange is very huge, complex, and volatile in nature. Prediction task for this market is highly challenging as the data is highly chaotic, volatile and noisy. In this work Artificial Neural Networks (ANN), Functional Link Artificial Neural Network (FLANN), Extreme Learning Machine (ELM) are the models used to predict the price. Simple Moving Average (SMA), Stochastic Oscillator, Exponential Moving Average (EMA), Momentum, Moving Average Convergence Divergence (MACD), Average True Range (ATR), Relative Strength Index (RSI), are different technical indicators used by economists to gain an insight into the market and predict the exchange rate of currency. Generally technical indicators are calculated from price, open price, low price, high price, change percentage. The proposed network is optimized by Genetic algorithm (GA), Differential Evolution (DE), Particle Swarm Optimization (PSO), Water Cycle Algorithm (WCA). The dataset collected for this experiment comprises of 4000 days of past currency exchange rates of the two currency pairs that is USA Dollar (USD) to Indian Rupees (INR) (USDINR) and Soudi Arabia Riyal (SAR) to INR (SARINR). The proposed datasets are segregated into many parts and each part is trained individually. Optimization techniques such as GA, DE, PSO and WCA deployed in segregated datasets as well as the whole dataset. The experimental result shows that the segregated WCA is giving the better result when WCA applied on the whole dataset. The ELM and WCA with segregated dataset produces better result than other models what experimented in this work.
Keywords: Forex; Water Cycle Algorithm (WCA); Differential Evolution (DE); particle Swarm Optimization (PSO); Genetic algorithm (GA); Extreme Learning Machine (ELM).

Scope of the Article: Machine Learning