Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: stacked generalization

Sortuj według:

Ogranicz wyniki do:

On Certain Limitations of Recursive Representation Model

Jastrzębski S., Sieradzki I.

Schedae Informaticae

2016

Vol. 25

37--47

There is a strong research eort towards developing models that can achieve state-of-the-art results without sacricing interpretability and simplicity. One of such is recently proposed Recursive Random Support Vector Machine (R2SVM) model, which is composed of stacked linear models. R2SVM was reported to learn deep representations outperforming many strong classi-ers like Deep Convolutional Neural Network. In this paper we try to analyze it both from theoretical and empirical perspective and show its important limitations. Analysis of similar model Deep Representation Extreme Learning Machine (DrELM) is also included. It is concluded that models in its current form achieves lower accuracy scores than Support Vector Machine with Radial Basis Function kernel.

A stacked generalization framework for the prediction of corporate acquisitions

Tartari E., Doumpos M., Baourakis G., Zopounidis C.

Foundations of Computing and Decision Sciences

2003

Vol. 28, No. 1

41-61

Over the past decade the number of corporate acquisitions has increased rapidly worldwide. This has been mainly due to strategic reasons, since acquisitions play a prominent role in corporate growth. The prediction of acquisitions is of major interest to stockholders, investors, creditors and generally to anyone who has established a relationship with the acquired and non-acquired firm. Most of the previous studies on the prediction of corporate acquisitions have focused on the selection of an appropriate methodology to develop a predictive model and the comparison with other techniques to investigate the relative efficiency of the methods. On the contrary, this study proposes the combination of different methods in a stacked generalization context. Stacked generalization is a general framework for combining different classification models into an aggregate estimation which is expected to perform better than the individual models. This approach is employed to combine models for predicting corporate acquisitions which are developed through different methods into a combined model. Four methods are considered, namely linear discriminant analysis, probabilistic neural networks, rough set theory and the UTADIS multicriteria decision aid method. An application of the proposed stacked generalization approach is presented involving a sample of 96 UK firms.