/ZS-Data-Science-Challenge

A Data science challenge - "Mekktronix Sales Forecasting" organised by ZS through Hackerearth platform. Rank: 223 out of 4743.

Primary LanguageJupyter Notebook

ZS_Data_Science_Challenge

  • This is the challenge organised by Hacker earth.

This problem involves:

  • Handle data to pre-process and generate relevant features and insights for modeling.
  • Knowledge of machine learning/statistics.
  • Identifying the right technique to approach the solution.
  • Different techniques to improve the accuracy.

Challenge Time:

  • 65 hours

Rank:

  • 223 out of 4743 participants

Language:

  • Python

Files:

  • Mekktronix_Sales_Forecasting Notebook - Contains the implementation of different algorithms that were applied on the preprocessed data.
  • Arima algorithm - Contains the implementation of Arima on the preprocessed data.
  • Presenattion - Contains the approach towards the problem and the details of the challenge.
  • Approach Rmd file - Approach file(Rmd version of Presentation).
  • SMAPE file - This file contains the SMAPE score on train data for different algorithms.
  • Predicted file - This csv file contains the predictions that were predicted on test file.(The values are achieved using xgboost algorithm)

Developer: