I believe a great part of this platform is learning from others.We will really appreciate if the winners can willingly share their solutionsThank you

Hello everyone, Firstly, congrats for the winners and thanks to Zindi for offering this competition. I think I want to give a brief description about my solutio... See more

Just trust your private LB. When I saw the public LB score, I was so determined that I ignored my private LB. I wish I could rely on my CV. I'm excited to see w... See more

Let's say our final prediction is preds_df, is it possible to perform the following operation in the prediction file? path = '' DF = pd.concat([ pd.read_csv... See more

How are you handling merchants with multiple categories?Eg: train.loc[train['MERCHANT_NAME']=="KAPS PARKING"]has "Transport & Fuel", "Bills & Fees",... See more

Hi all, I believe feature engineering plays a huge part in this competition.I am looking for a teammate that has a good model, and the goal is to combine our fe... See more

Any tips on how one can use the extra unlabelled data?

This is my code solution to the competition, thanks @zindi for a great competition and congrats to whoever wins the competition(I just won't have enough time t... See more

What is your current best single model?for me :model : catboost fold : 5 LB : 1.33441214352313 CV : 1.2544563969437823

my cross-validation scores are pretty far from the leaderboard scores! Any help, please.

dears,i have issue when i submit file , issue is missing id for predection but when i open file i found the id inside the sheet with prediction .kindly advice... See more

@ZIndi Is using WEB SCRAPING okay?

Does anyone please know how to solve this issue; Thank you!ValueError: y_true and y_pred contain different number of classes 8, 11. Please provide the true labe... See more

I have a set of NumPy arrays in a list and some of the NumPy arrays don't have the same shape as their other members. I will like to pad them with zeros. Does a... See more

***Corection, I confused the metric with another competiton***But there's still an issue on predictions made with no technique at all getting better scores than... See more

Funny thing - if you submit submission file, filled with any random number (lets say 0 or 1) you will get score near 2.56, which is much better then bencmark :)... See more

The purpose of this competition is to predict the payment type for new transactions. The popular way to split the labled set is to separate the train set and te... See more

dears, all predicted model from my side , has low accuracy and models not predict all classes, more than 3 classes not predicted , is this natural ?

What was your Baseline Score without feature Engineering?

I observe that education and rental mortgage have two values each in the training dataset. I can not fit my data with stratified or K-fold cross-validation. Wha... See more

is this competition about multiclassification or multilablel classification or both?