Deep sequence modelling of Data Envelopment Analysis - Based performance measurements
Ακολουθιακά μοντέλα βαθιάς μάθησης για την εκμάθηση μέτρων αποδοτικότητας βασισμένα στην Περιβάλλουσα Ανάλυση Δεδομένων
View/
Keywords
Machine learning ; DEA ; Sequence-to-Sequence ; Deep learning ; Neural networksAbstract
In this section of the study, a summary of the project and its key areas are presented. The following study is focusing on exploring and proposing different algorithms in the Artificial Intelligence scope that can manipulate and handle the data transformation from a dataset of a Data Envelopment Analysis efficiency scores’ prediction problem, to conclude what are the key elements that make up this score and later to determine whether this kind of scores prediction is possible or not. The project starts with some introduction on the area of the Artificial Intelligence field of Computer Science, continues with a brief presentation of the DEA operations research technique and similar studies that have tried to combine the DEA research together with the Machine Learning sub-field from Artificial Intelligence.
Concluding, having presented the approach on the thinking behind the creation of the model to resolve this Machine Learning problem, the benchmarking of the performance of the model is presented, following by a brief description of the conclusions that were reached from the simulation of this algorithm execution.