Primary competition visual

IndabaX Eswatini: VAT File & Pay Predictive Analytics Challenge by IndabaX

40 000 SZL
15 days left
Machine Learning
Feature Engineering
Prediction
63 joined
20 active
Starti
25 Sep 26
Closei
11 Oct 26
Reveali
12 Oct 26
About

The data covers VAT returns filed with the Eswatini Revenue Service. Each row is one VAT return, meaning one taxpayer for one filing period.

Train.csv has two target columns:

Column	Classes	                                 Meaning
FILING	ON_TIME, LATE_1_7, LATE_8_30, LATE_31PLUS	Days after the due date that the return was filed
PAYMENT	NONE_DUE, ON_TIME, LATE_1_30, LATE_31PLUS	NONE_DUE means nothing was owed (a nil or refund return); 
                                                otherwise, days after the due date that payment was made

A small number of FILING and PAYMENT values in Train are blank because the outcome was not yet known on 1 July 2025.

Columns in both Train and Test

  • ID: unique return identifier
  • Taxpayer_ID: taxpayer identifier, the same across Train and Test
  • Class, Division, Sector: nature of business
  • Filing Frequency: Monthly, Quarterly or Annually
  • Taxpayer Type: individual or non-individual
  • Segment: taxpayer size segment
  • Organization Type: type of organisation
  • Start Date, End Date: the tax period
  • Obligation Due Date: the date the return and payment were due

Columns in Train only

  • Return figures: I_O_Ratio, Zero_rated_IO, Exempt_IO, Effective_Output, Zero_Rated_Share, Exempt_Shares, VAT Refundable/Payable, Refund_Position, NIL_FILER
  • Dates: Filing date, Payment Date

These only exist once a return has been filed, so they are not available for Test. Use them to describe a taxpayer's earlier returns, for example their previous VAT payable or how often they file nil returns.

Submit one row per Test ID, with a probability between 0 and 1 in each of these eight columns:

FILING_ON_TIME, FILING_LATE_1_7, FILING_LATE_8_30, FILING_LATE_31PLUS, PAYMENT_NONE_DUE, PAYMENT_ON_TIME, PAYMENT_LATE_1_30, PAYMENT_LATE_31PLUS
Files
Description
Files
This is the test data to which you will apply your model. This is the dataset against which your model will be evaluated.
This file shows the format of your submission file.
This is the data required for training your model.