A Comparison between Nature-Inspired and Machine Learning Approaches to Detecting Trend Reversals in Financial Time Series.

摘要: Detection of turning points is a critical task for financial forecasting applications. This chapter proposes comparison between two different classification approaches on such problem. Nature-Inspired methodologies are attracting growing interest due to their ability cope with complex tasks like classification, forecasting, and anomaly detection problems. A swarm intelligence algorithm, namely Particle Swarm Optimization (PSO), an artificial immune system Negative Selection (NS), have been applied the detecting points, modeled as Anomaly (AD) Particular attention has also given choice features considered inputs classifiers, significant impact they may overall accuracy approach. In this work, starting from set eight input features, feature selection carried out by means greedy hill climbing in order analyze incidence reduction global The performances obtained compared other traditional machine learning techniques implemented WEKA both methods found give interesting results respect techniques.