Period Detection and Future Trend Prediction Using Machine Learning Techniques
Haoye Lu, Anand Srinivasan, Amiya R. Nayak · 2018
Period detection and trend prediction algorithms have widely ranged applications in many areas. Data involving periodic properties are omnipresent. However, while many general prediction methods are proposed, the prediction algorithms related to periodic data are hardly discussed. Also, period detection methods are still limited to the applications of autocorrelation functions. In this paper, we propose an algorithm, using learning automata techniques, to predict future trend and detect period. Given a repeating sequence, our method can automatically find its period and make predictions on its future values. To the best of our knowledge, this is the first algorithm that can automatically find the period of the inputs and further use it to predict future trend. The theoretical analysis and simulation results are also discussed in this paper.