Hi everyone,

I am currently working on a Multi-Criteria Decision Making (MCDM) based classification model. Recently, I submitted a paper on MCDM-based ordinal classification. The reviews I received were primarily focused on machine learning and deep learning methodologies. Here are some of the key comments I received:

  • The candidate needs to improve their awareness of the domain. So far, the candidate has only worked with a subset of classical machine learning methods applicable to this type of problem. The candidate should learn advanced methods for modeling, such as Feedforward Neural Networks (FFN), Convolutional Neural Networks (CNN) for image data, and Recurrent Neural Networks (RNN) and transformers for Electronic Health Records (EHR) data. Additionally, understanding interpretability methods, including gradient-based methods, DeepLIFT, Shapley values, LIME, and Integrated Gradients, is crucial.
  • The candidate should explore multimodal data modeling and model uncertainty quantification. Understanding model uncertainty will help the candidate assess the reliability of their model predictions. This is a well-researched area that will be beneficial for the candidate's work.
  • The candidate needs to read more literature and visit hospitals to better understand the challenges and significance of the problem they are trying to solve.
  • Can anyone help me defend my work in response to these reviews? My research is not focused on the suggested areas, but I received this feedback. I would appreciate any guidance on how to address these comments effectively.

    Thank You .

    More Soumita Guria's questions See All
    Similar questions and discussions