The Application Of Data Mining And Machine Learning In Astronomy

This article describes the current state of data mining and machine learning in astronomy.From the perspective of researchers in this field,data mining may have some mixed implications.If used properly,it can be a powerful method that has the potential to take full advantage of the exponentially increasing amount of available data,thereby promising huge scientific advances.However,if misused,it may just be a black box application of a complex computing algorithm,which may provide little physical insight and provide questionable results.This article outlines the entire data mining process from data collection to interpretation of results,covering common machine learning algorithms.It mainly explores the application of data mining and machine learning in astronomy,specifically using automatic clustering algorithms combined with planetary spectral data.Classify the photometric data of the planets,review and collect relevant data,organize and summarize according to the collected data,and mainly introduce the principle of four machine learning algorithms and data mining.And then use an automatic clustering algorithm to analyze and process the observed planetary spectral data and classify it,and analyze the classification results in detail to determine whether it has achieved the expected results.The biggest feature of this paper is that it not only requires the classification model to have good generalization performance,but also improves the selection of data to obtain better classification results.We use the z-score method to obtain different data sets by setting different z-score sets,which makes our data more centralized.We roughly compared a suitable set of scores,which makes the classification model better.