Thesis tips

The author of the paper tip:

1、 Reviewers don't spend much time reading papers , After all, everyone is busy . The core contribution Be sure to write clearly .

2、 be based on 1, Workload is the top priority .

3、 Don't have grammar problems , It can be used grammarly Check .

4、 Do not have undefined variables or ambiguous variables , If there are some nouns , Be sure to quote or explain it , Explain its relationship with the paper .

5、 Don't use some abnormal nouns .

6、 The effect should exceed baseline.

7、 stay introduction It's better to add a picture to the part , A good paper hopes to pass introduction You can know your job . chart > surface > written words > Algorithm .

rebuttal ：

1、 Reviewers generally don't change their minds easily , It won't take long to see rebuttal.

2、 Don't just prove , Even if he gave it to you reject, It should be justified .

3、 be based on 1,rebuttal Based on experiments rebuttal, Don't just use theory , If you don't want to add experiments, you should also add experiments to explain .

4、 Answer any question clearly , In particular, it involves novelty The problem of .（ Don't let review Go directly to the article xx Duan Xun , At least explain what has been modified in the paragraphs and sentences ）

5、 Encounter biased reviewers , The key is to show others .

summary ：

1、 The reviewer of your manuscript is not necessarily in your field , stay introduction Be sure to introduce the importance of your field .

2、 To understand community Some accepted principles of , Don't make any assumption. for example resnet, This quotation and introduction will introduce , The layman cannot understand what this is .（resnet It is an extreme example ）. There are research hypotheses in the theoretical part , Cite recent papers to support your argument , And these papers should preferably come from core journals or top Journal .

3、motivation Logical chain of （ The method must be right , Demonstrable , I can't say I feel xx It is feasible. , Then it's really feasible ）, A small experiment of hyperparameters , Visualization results 、 The analysis process . The experiment must support the existence of your thesis . All experiments should be related to your novelty.