UNDERSTANDING DATA SCIENCE WITH “PUBG”- 1
“Winner Winner Chicken Dinner”
In recent times a game that is almost played across geographies with a theme is best to think about courage, strategies and team work that will make a chicken dinner . I always believe that we read many concepts but only some aspects which we relate to things we like will make a difference in practical implementation . 1 in 10 data science enthusiasts do high rated certified courses and video sources but has always been a question in my LinkedIn stating “ I don’t know how to do practical data science, seems I have done many courses but still not able to figure out what exactly correlate to real world” . I’ve also been in the same stage at my career’s start but when i got to understand slowly, it seemed really interesting to work with . “The actual concepts, may it be text book or videos or courses, they always potrait the author’s point of view. So we just have to add our way of understanding with the original concepts to reach a saturated point of understanding”. So here iam to showcase my understanding in publically loved themes so that enthusiasts can better relate the concepts in correct way.
The Rule Of PUBG :
- 100 people join in a flight and choose to land anywhere in a map
- Teams or individuals will loot in all places and get ready for the war
- Makes sure if the guns and medical is in stock
- Makes sure he is inside zone
- Fights the enemies and wins a chicken dinner
Strategies for survival :
- Keep calm and keep up the survival ratio
- Get kills and increase kill ratio
Way to approach Data :
“Form up on me”
The strategic Survival in pubg is to be followed keenly with data. Let’s make our relavency clear by specifying terms :
- The Bag : The variables you choose for a particular problem statement . The bag in pubg has 3 levels and so should be your way of choosing the container for the data . At first when you enter the game, you try to pick any bag may it be lv.1 or anything, and you will try to fit only necessary items — so should be your way of choosing variables in your dataset . Consider the bag ( the dataset container ) to be a table with 'n' columns as variables and rows to be the data for those lables.
- The Loot : The items inside or outside necessary for the battle will be made available . These are things like machine guns, snipers, shot guns , medikit, extended mag, suppressor, pan, bullets, scope etc and other necessary accessories . These are nothing but data ie, the rows of the container . How much you pick and how many duplicates you drop makes perfect sense when you don’t have space in the bag . So you carefully tend to choose with the quantity or amount of things that you carry .
- The Scope : This is an extra fitting that should be added to the gun so as to see a long shot and take a head shot . This varies with 3x,4x,6x,8x with which we can take a shot on enemies who are away from us . This activity refer by taking a deep look at the data and making it clean and ready for us to feed to the model. So please do put an 8x with sniper to carefully look into the data and just shoot out the duplicates, missing values (can also be replaced) , eliminating outliers and make the data clean . Also remember to use RedDot for AKM since it shake a lot.
- The Zone : In the game, if the player cross a particular regional area, he might tend to loose his health . So here , we shall consider the zone to be “Train-Test-Split” . So you should be very much clear in the split ratio which is nothing but the amount of data that has to be split so as to train the model and test the model . So if a particular amount exceed invariably in the split, we may end up with overfit or underfit of data which might result in bad outcomes. Be careful that you are always inside zone.
These are some of the basic ideas to start with .If you like the theme and want me to continue, please do follow and give your support by clapps and if you wish to join our group of experts to make a difference and solve real world problems please do leave your details in comment section . If you want me to explain in any other themes, please do contact or leave in comment.
Contact Details —
Mail : technobuff.ml.ai@gmail.com
Member contact :
https://www.linkedin.com/in/aishwarya-balachandran-3a8013114/
Instagram Id : deep_ml5.ai
Telegram handle : @machineLearningAndPython
