<style> .reveal { font-size: 28px; } </style> # Team $TEAM_NAME Template for presentation team's solution. Example: https://hackmd.io/s/slide-example [TOC] --- This is a general presentation - it may not fit to any competition. Please use it as a template, ignoring irrelevant questions and adding whatever information you'd like to add (memes included). You can sign up in HackMD, copy the this markdown and create your note. To show it in Slides mode, you need to click "Share" button, and change "Edit mode" to "Slide mode". --- ## Data processing :spiral_note_pad: - How do you process the data? Which features were considered? - Do you find additional data? - What improved the score and what worsened? Why? - What should worked, but didn't. Why? --- ## Train models :hourglass_flowing_sand: - Which libraries and algorithms did you use? - How did you tune the models? - If you used neural networks which architecture did you try? - What improved the score and what worsened? Why? --- ## Validation of your solutuion :thinking_face: - How did you measure the quality of solutions? - How did you choose the best solution? --- ## Work in teams :family: - How did you organize work in team? - How did you mix yours solutions? - If one person processed the data and another trained the models, how did you share information? - What was helpful and what wasn't? - What would be done differently in next time? --- ## Implementation :clapper: - Was implementation maked locally or on server? - Which tools were used? - Which guidelines were used to keep code and work in order? - Which tests were written and what exactly was tested? --- ## Key insights :roller_coaster: Action points, which you wrote down - 1 - 2 - 3 --- And also, general questions for all topics. I will list them here to don't repeat every time :smile_cat: - Which interesting ideas did you have? - What did you want to try, but couldn't/didn't have time? - What did work good? - What didn't work good, but you thought what it should work good? - Which difficulties did you face? - Which useful conclusions can be drawn to remember for the future? - What would save you time, if you knew about it in advance? - Something that you want to share with others:+1: --- ## Sharing your work 1. Publish your solution as a Github repo or, if it's small enought, as Kaggle kernels notebook :octopus: 2. Publish a (brief) write-up, describing your solution. You can use this presentation - just add verbosity and context to turn it to something readable. Suggested places: kaggle discussions, medium, dev.to or telegram channel if you have one. You need to publish it at least once somewhere, but other team members are encouraged to repost that to increase your visibility and get some support and feedback from your friends and colleagues :hugging_face: 3. Post links to your repo and write-up to Kaggle discussions if that makes sense. :loudspeaker: Add links to your code and write-up to this presentation.