amherag / draft-papers Goto Github PK

writing is bad
** se lo paso a mi amigo de canadá jajaja. aunque estaría bien tener a alguien dedicado a esto. además se podría referir a la estructura del paper (revisar coherencia y cohesión)

literature research. Apparently the authors have not done a good literature research.
** usar referencias más recientes y que sean artículos del mismo TFS de preferencia

However, the theoretical results should be improved greatly, if the manuscript can be qualified for being published in TFS.

The main results are comparatively weak and also, the simulations are too much.
** Cambiar el dataset, talvez el algoritmo, quitar simulaciones

Authors did not discuss the datasets very clear and how to do the data preprocessing
** Cambiar los datasets mejor?

Benchmark technology to compare with: the authors compare their work with Brown's work. Is Brown's work a benchmark technology in this domain? In addition, no other benchmark technologies have been compared with.
** Usar mackey-glass y/u otros datasets mejor.

De acuerdo, este es el paso importante. Un benchmark conocido de Series de Tiempo y además otro de stock exchange. El punto es que haya muchas publicaciones que reporten resultados y técnicas.

Issues #12 #13 y están relacionados.

Authors should use the statistical methods to support their findings with other algorithms, even to evaluate the normal distributions of their experimental sample.

I appreciate if authors can adopt more popular stock index or international stock market as their experimental design and datasets to prove the proposed methodology has significant prediction performance in different filed.

The relative researches about time series or price prediction model are very popular and I suggest authors should extend their literature reviews about this field.

Performance evaluation. The majority of performance evaluation of this work has not been compared with others. I do not understand how the authors think this proposed method is superior.

Section II shows a mass of related work in reference, which is better to be stated in the introduction and cannot be viewed as a separate charter.

Finally, they should draw their conclusions in the Section 5. In addition, authors should discuss the future works and research limitations more clear.

In is paper, authors did not explain very clearly about how to integrate the FIS, MAS, and GP, besides, how they decide to setup these parameters of this model based on author's experiments?

In the Introduction section, authors should explain or discuss why they adopt FIS, MAS and GP to construct the hybrid DSS model. What's the advantage or the novel approach of this proposed model?