Bank Transaction Predictor Data Science Sprint








    Status: ‌Cancelled failed review
    Show Deadlinesicon-arrow-up

    Challenge Overview

    Hey there Topcoders! Just wanted everyone to know that this is a data science challenge, but since the client would like this created in Python, we decided to run it as a code challenge.
    Our client would like to run a single data science sprint to generate a proof of concept (POC) which predicts future transaction details. You will be creating this program from scratch; we have provided data and resources.


    Final Submission Guidelines

    The goal is to predict next payment in each separate stream of payments that might be more than 30 days ahead. The POC is both finding lines of regular payments, and then predicting the next payment.  The input data is a set of financial transactions. From the input data the POC will be focused on predicting three main points of each transaction:
    1. The date of the next transaction
    2. The amount of the next transaction
    3. Each separate transaction if multiple transactions occur to the same entity in a single month

    Here are significant dates for you to consider:
    • Holidays in Spain-  predict how many labor days (work days) until the next payment.
    • 3 main kinds of income in Spain and their pay dates
      • Employee: paid at end of month usually, though occasionally on 1st of month. Direct deposits get messed up
      • 10th of month is unemployment benefits
      • 25th is pension plan for those who are retired
      • 5th is for pregnant mothers who receive income through the state
    The primary goal of each prediction is to be within a margin of error <1.5 days and <15€

    Out of Scope

    The following is not in scope for this challenge:
    • Unit tests
    • Security testing
    • No UI will be developed

    Tech Stack


    Submission Guidelines

    • Deploy your submission to an AWS environment. Please include a link in your README
    • Please submit a .zip file with your source code
    • Your README should have directions to install and run your app, as well as any configuration information.

    Reliability Rating and Bonus

    For challenges that have a reliability bonus, the bonus depends on the reliability rating at the moment of registration for that project. A participant with no previous projects is considered to have no reliability rating, and therefore gets no bonus. Reliability bonus does not apply to Digital Run winnings. Since reliability rating is based on the past 15 projects, it can only have 15 discrete values.
    Read more.


    Final Review:

    Community Review Board


    User Sign-Off


    Review Scorecard