I am cofused which one should be our Feature columns and which one should be target?
Can someone help please? I am new to compition
Hello @Bhavika, please refer to this community discussion thread by @johnowhitaker , you will have great clarity where and how to start
@johnowhitaker has given great note on how to start with notebook. I am also new to competitions on satellite image data processing and his notebook has helped me a lot to get start with. This is what I have understood so for,
- you read the image data by area and location.
- extract images/tiles/chunks from the large .tif image files with the help of geodata.
- at the same time convert the same image into image mask. Image masks are something like black and white image. E.g. in this competition roof tops will appear white i.e. pixel values 0’s rest will be black with pixel values as 255.
- these image mask will act as your targets and actual images as your raw inputs to your model.
hope this helps. happy learning…
Hello @Bhavika, I don’t really understand the confusion, but as far as I understand you want to know the following:
- the training features can be found in the Training Data Features file, dengue_features_train.csv
- the training labels can be found in the column ‘total_cases’ of the Training Data Labels file, dengue_labels_train.csv
- You then try to find the the total cases for the features given in the Test Data Features file, dengue_features_test.csv
- in the end you add your results to the ‘total_cases’ column of the submission_format.csv fiel and upload the result.
I hope I understood your question well??