when I submit a submission file(even the SampleSubmission.csv), It turns out that "Error loading data".why would this happen?

Hello, is it possible to get permission to use the dataset for this competition for a dissertation?

I want to congratulate all winners and thanks organizers for this competitionmy solution is a blend of 2 types of models 20 binary classifications Multyclassifi... See more

You can find my solution in github: https://github.com/dead-mazai/Zimnat_Insurance_top-9Don't judge me harshly, I'm just a second-year student)

Just curious, but was there a pre-existing benchmark to begin with? Just wanted to see if it was possible to have a decent gauge of my newbie skills

please find my solution in github: https://github.com/berndAllmendinger/Zimnat-Insurance-Recommendation-Challenge.it is a singel lightgbm model with few engenie... See more

Hi. Waiting for the winners solutions. Good job.

Congratulations for boys from top-3, ur so good, i wish to see what u have done in notebooks. It will be good experience for all if you give us these!And bad ne... See more

can anyone suggest me which one was the best algorthim works for this data

Hi, is it practical sharing winners code here or it is restricted ?

What is the distribution of data for public and test?

Hi. Is there any assumption that there is relationship between observations, say several observations belong to a family?

Hi,1) it normal to use test data with ones in train process, assuming that rows are already known, and will add some info in training process.2) how handle valu... See more

@ZindiAdmins, With the above statement, is it safe to assume that we CANNOT calibrate our probabilities by any manual process. The ones that are predicted can b... See more

There are very few positive samples (ie. 5) against GVSR column how can we handle this? Even if we use SMOTE it might overfit or underfit the data.

For the test file, we have to make a prediction for each product (0,1) for each customer. But I am unable to understand the format of the submission file. For e... See more

How did anyone manage to reshape the data to fit the submission format

Can someone please confirm or tell us the most influential/important feature in predicting target feature?

I have tried approaching this problem using collaborative filtering but it gave me a really low score. I read through the comments that the best method here is... See more

Good day people,I would like to pick your beautiful brains on the treatment of the disparity on occupation codes. It appears there are 233 unique values on trai... See more

Hi everyone.is it possible to count how many 'ONE' is missing from the test data?

hi there. what does meen merital_status = 'M' when birth year is 2011? that is the user is already maried?

Which is the best way to prepare the data?1. Melt the products columns and make it a binary-classification problem.2. Duplicating the rows for all the products... See more

Catboost cannot reproduce the exact result (the metric floats all the time) on the GPU and it really seems to me that on the CPU the same (I checked only 1 time... See more

the output of my classifier after calling model.predict_proba() gives two columns probability that it is a 0 and probability that its a one how do I m... See more