Please ans the 15 question below according to this article... Deep learning is another emerging field in machine learning. It is still under exploration stage because the neural networks developed under it are crucial to understand machine learning or artificial intelligence. AI is used for decision-making and DDN are helpful in finding answers of the questions, which normally remained unanswered in the process. As a user if one is not satisfied with the answers the trust and reliability is questioned. The article is quite helpful for new researchers and potential students who want to pursue deep learning as a career. It is going to help them understand about the dimensions the methods used, their explanations and the way of evaluation, prospective research topic which can be aligned with the topic, the required considerations for the designing phase and the future possibility in the field (JAIR, 2022). The explanation stage is quite interesting as it helps to understand the models better and generalize them for better communication. Secondly the explanations help to predict the outcome from the models. These are in turn useful to understand about decisions and the factors on which it is dependent. The author explained different methods for describing DNNs namely: Model distillation: This method helps to understand the input-output behaviour so that decision can be discuss in detail and understood through input measures and output DNNs. Visualization method: A scientific visualization of an input's characteristics that have a significant impact on a DNN's output is used by visualization methods to convey an explanation. Intrinsic method: DNNs that were specifically designed to provide an explanation alongside their output are known as intrinsic methods. Intrinsically explainable deep networks can jointly optimize model performance and the form of the explanations produced as a result of their design (JAIR, 2022). There are variety of topics that are associated with explain ability of machine learning for instance learning mechanisms. To let users understand the semantic meaning of models this approach is used. Another consideration is model debugging which helps to identify any error related to data processing, training and architecture of the model. It contains the idea of neural stethoscopes which helps in investigating the presentation of network layers. When the judgement regarding DNN is of doubt then the concerns of attack and defence arises which means system becomes vulnerable to attacks and the input method is under radar. The concept of fairness is relevant and author highlighted the fact that biases need to be checked before entering information in the system so that no mistreatment with the information can happen. Normally fairness issues arise in pre-processing methods, in-process methods and post-process methods (JAIR, 2022). The explanations holds important for users who are going to work on designing part because it is required to let user understand the purpose, the workings and satisfy his queries for the same. The author explains about end user in detail so the reader can relate it well to the topic. The DNN decisions and their impact the speed with which the decision is taken. The future directions shown by the article in the form of explanations which can be user friendly and unified in terms of AI so that machine learning can be easily understood.

After you read the article you've chosen to summarize, answer the following questions:

1. The primary audience for this article is

2. Other potential users are

3. The purpose of the article is

4. I will use the document to

5. As an audience, I know (background knowledge about the subject)

6. As an audience, I need to know

7. Here are some questions I have about the article

8. My attitude toward the topic is

9. My objections are

10. Those most affected by the article are

11. People that might be alienated by this article are

12. The kinds of information in the article most important to me are

13. The article is organized in this way

14. The tone of the article is

15. The article's intended effect is