作者: José L. Pastrana , Rafael E. Reigal , Verónica Morales-Sánchez , Juan P. Morillo-Baro , Rocío Juárez-Ruiz de Mier
关键词:
摘要: Data mining is seen as a set of techniques and technologies allowing to extract, automatically or semi-automatically, lot useful information, models, tendencies from big data. Techniques like "clustering," "classification," "association," "regression"; statistics Bayesian calculations; intelligent artificial algorithms neural networks will be used extract patterns data, the main goal achieve those explain predict their behavior. So, data are source that becomes relevant information. Research gathered numbers (quantitative data) well symbolic values (qualitative data). Useful knowledge extracted (mined) huge amount Such kind allow setting relationships among attributes sets, clustering similar classifying attribute relationships, showing information could hidden lost in vast quantity when not used. Combination quantitative qualitative essence mixed methods: on one hand, coherent integration result interpretation starting separate analysis, other making transformation 1 vice versa. A study developed shows how can very interesting complement methods, because such work with together, obtaining numeric analysis based probability calculation transforming into using discretization techniques. As case, Psychological Inventory Sports Performance (IPED) has been mined decision trees have order check any "Self-confidence" (AC), "Negative Coping Control" (CAN), "Attention (CAT), "Visuoimaginative (CVI), "Motivational Level" (NM), "Positive (CAP), "Attitudinal (CACT) factors against gender age athletes. These also for future predictions assumptions.