🏆 PokeMon Webinar Challenge PokeMon's Simple Classification Challenge Solve Challenge & Grab a Deep Learning T-Shirt! You are going on a trip to Dudhsagar Falls, Goa. To reach the falls you have to cross a huge forest in Collem, so you thought of entering the forest. While you are on your route you spotted multiple Pokemons like Bulbasaur, Pikachu, and Charmender. After coming back, you thought of making a PokeDex - a Pokemon Detector Device which classifies these Pokemons. Write an algorithm that classifies these PokeMon's.
About Dataset There are two folders Test and Train.
Train: The train folder consists of a folder named Images which contains the images of three types of pokemon and a csv file which maps the images in the image folder to the name of that pokemon, The csv contains the field ImageId which is nothing but unique name given to each image and other field is NameOfPokemon which gives the pokemon name associated with that image.
Test: Test also contains an image folder on which your model will be tested on and also there is a test csv file which only contains the name or ImageId of the image and gives the order in which each image is to be labelled .
Submission: This file will be a csv file that will be submitted by you it will contain the same fields as that of sample submission the order of the ImageId should match that of sample submission or test csv.
The three categories of Pokemon to be classified are Pikachu,Bulbasaur and Charmander.
Submission Instructions Write entire code from scratch, don't forget to resize images to a smaller size to increase efficiency, and normalise pixel values between 0-1.
You can use any algorithm of your choice (KNN, Logistic Regression,SVM, Neural Network etc)
Use of sci-kit learn is not allowed
Submit your predictions as .csv file. Make sure you have proper headers in the file. -Make sure the predictions are in same format as train csv i.e the name of Pokemon along with cases must be checked before submission.
Winners will be announced on 11th July on Coding Blocks Youtube Channel.