Scribd Logo
Upload

Welcome to Scribd!

  • 16 views

    Miniproject Hints

    Uploaded by

    Satheesh Manikanta
    The document provides guidance on how to write a good mini-project report by avoiding common mistakes. It recommends including a task description, summary of methods and results, detailed methods section, results with both positive and negative findings, discussion of lessons learned, and references list. Basic report elements like figure captions, symbol definitions, and citation format should follow standards. Crucially, experiments must be described in detail enough to allow replication. The writing style should be formal and avoid colloquialisms, and only cite references actually read. Methods should be honestly described, including limitations in understanding. Cross-validation is emphasized as key to modeling.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    Miniproject Hints

    Uploaded by

    Satheesh Manikanta
    16 views2 pages
    The document provides guidance on how to write a good mini-project report by avoiding common mistakes. It recommends including a task description, summary of methods and results, detailed methods section, results with both positive and negative findings, discussion of lessons learned, and references list. Basic report elements like figure captions, symbol definitions, and citation format should follow standards. Crucially, experiments must be described in detail enough to allow replication. The writing style should be formal and avoid colloquialisms, and only cite references actually read. Methods should be honestly described, including limitations in understanding. Cross-validation is emphasized as key to modeling.

    Original Description:

    Its all about miniproject

    Original Title

    MINIPROJECT HINTS

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document provides guidance on how to write a good mini-project report by avoiding common mistakes. It recommends including a task description, summary of methods and results, detailed methods section, results with both positive and negative findings, discussion of lessons learned, and references list. Basic report elements like figure captions, symbol definitions, and citation format should follow standards. Crucially, experiments must be described in detail enough to allow replication. The writing style should be formal and avoid colloquialisms, and only cite references actually read. Methods should be honestly described, including limitations in understanding. Cross-validation is emphasized as key to modeling.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    16 views2 pages

    Miniproject Hints

    Uploaded by

    Satheesh Manikanta
    The document provides guidance on how to write a good mini-project report by avoiding common mistakes. It recommends including a task description, summary of methods and results, detailed methods section, results with both positive and negative findings, discussion of lessons learned, and references list. Basic report elements like figure captions, symbol definitions, and citation format should follow standards. Crucially, experiments must be described in detail enough to allow replication. The writing style should be formal and avoid colloquialisms, and only cite references actually read. Methods should be honestly described, including limitations in understanding. Cross-validation is emphasized as key to modeling.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 2

    Howtowriteagoodminiprojectreport,orrather,howtoavoidfrequentlymadeblunders Herbert Jaeger, November 1, 2009

    Overallstructure A recommendable structuring is the following: 1. Task description (a summary of given task, plus summary of motivation: what specific difficulties will have to be solved and what should be learnt from miniproject) X. You do not need to write a scientific-paper style introduction section with an outline of scientific background, related work or research motivation 2. Summary of what you did: sketch of methods used, results obtained, lesson learnt 3. Methods: description of what you did, in enough detail and formal precision to allow replication! 4. Results: documentation of results (also negative ones, if interesting) Y. You do not need to document your code in the report that should be done within the code. 5. Discussion: what you learnt, what you would do more/better if you had more time/methods available 6. References list.

    Basichygiene All figures need a caption and number, and need to be referenced from within text by their number. All symbols in figure, and axes, must be explained in caption or in text. All symbols used in text or equations need an explanation, either in plain English or a formal definition. For any non-self-generated graphics, give source For any non-self-generated math formula, give source Use a uniform and standard and complete format for your references. Check any article from a good scientific journal for a model, and adopt it. Different journals have different citation styles choose one that you like and stick to it.

    Reallyimportant Describe your experiments in enough detail to allow replication. A too superficial treatment is the strongest and most commonly encountered reason for downgrading reports.

    Style Avoid colloquialism and superlatives. Words like "sort of", "huge", "tremendously" have no place in scientific writing. Avoid global judgements of the kind "the most famous...", "the most widely used..." because you don't actually know this! Statements of this sort (expressed in proper terms) are only admissible for authors that have a long experience and a good standing in a field. In a student's report, they sound presumptuous.

    Don'tpretendtoknowwhatyoudon'tknow Only cite references that you have actually read. I frequently find cites of papers that are clearly too advanced or too classical (and hard to get) to believe that you read them and feel duped. Bad for the grade! Only write down equations that you understand. I frequently find equations taken out the the hardly understood context of some paper, with not all symbols explained, and not well connected to the rest of the report. Bad for the grade. If you want to use a method that is difficult (like SVNs) but for which ready-made program packages exist which you want to use you are free to use them, but you should honestly explain the extent to which you understood them. Engineers working "out there" often use methods they don't understand, no-one blames them if the results are good: but if they are not good, it's likely that they didn't use the methods properly. If they (or you) are aware of this, and overtly describe the limitations of your understanding and the details of how you did employ the methods, no reason for blame.

    Ageneralmachinelearninghint Cross-validation is the key to really most if not all modeling, employ it routinely. Cross-validation is more than splitting the training data into a train and validation set it has to be done repeatedly, going over all "folds"!

    You might also like