Peer-review-feedback-notes-lecture12.py

# Lecture 12 

# walk through of peer review -> 

      # Remember : kea name 
      # Henrik explains : group names will enhance Henriks exam orchestration
      # Methode : section (describe the methode framework ( no details included realted to the concrete data) 
      # VERY importatn!!! - SORUCES -> Dont write a lecture book -> and remeber sources -> and reference yann lecun ex. on a deep neural network
      # how to approach : ->> quick literature review before writing and article and scan and get an overview 
      # Only reference a source a dont descrube methods concretly itself 
      # INTRODUCTION :> is "setting the stage" -> how a motivation to read and reason
      # WRITE A MATHIMATICAL equation to describe the linear regression (henrik aknowledge as clear and concis)
      # WHITE SPACE 
      # two column should help fit the right amount of text for the study 
      # SENTENCES: beware of red
      # overfitting > big gap between train and test data results in overfitting
      # 
      # Explicit make a judgement before making any statment .> compare etc. 
      # We are allowed to have opinions in our conclusion at can make a judgement 


      # How do we know our result are good? How a reason to state that 

# My review 


      # good title -> 
      # gropp name -

      # PAGES !!! dont make to many exam >>>>!!!
      # source 
      # good strucutre 
      # discussion after conclusion 
      # hypothesis -> merthods is analyse to concrete to the problem 
      # to much analysis -> 
      # Meythode >
            #data cleaning -> 
                # data preperation ->
                  # feauture engineering 
                        # modelling 

      # first write the methode section : > i have made data preparation -> from using these sources -> Make thinngs repreducable 

              # METHODE :_> should precisly carve out which methodes are being used
    # START WITH THE METHODE SECTION
    # it should be including all methodes -> and not too short and is well supported by sources
    # New features -> explain why they are made and combining the feature -> WHY???? explaoimn !!!!!!!
    # miss explanation for what really goes on -> analysis is missing and understand why things are used 
      # show graf 
      # conclusion should be sharper - > direct adressing the problem
    # WE can use "we" we can use passive 

    # methode reference algorithms -> correlations -> pearsons DNN deep neural network 


      # MOST IMPORTANT BIBLIOGRAPHY _> 

      # PEER REVIEW 

                  # VISUALIZATION       
                  # GOOD GRAPHS ->