Ended on 16th Sep'22 03:00 PM (Coordinated Universal Time)

Juniper Networks Global AI Challenge

Predict hourly sales for cash registers across a retail chain

4.4k
Medium

Organized By:

Organized By:

Juniper

Rewards:

Rewards:

$5000

Challenge Starts

26 Aug 04:04 pm

Registration Ends

16 Sep 03:00 pm

Challenge Ends

16 Sep 03:00 pm

Announcement (8th Sept):

Hello Everyone! The deadline for the challenge has been extended by a week, and it will now end on 16th Sept. This gives all the competitors an additional weekend/week. We hope you make the most of it.



Further, below are a few important points that are relevant for eligibility for winning titles and prizes:

  • Your final model submission

    MUST NOT
    have experimental features. However, we would like to see how the performance changes when experimental features are added to your final model. All the top competitors' notebooks will be subject to manual evaluation. If any competitor violates this guideline, they will not be considered for the final leaderboard and prizes.

  • Be cautious of overfitting:
    Please note that ranking top on the public leaderboard may not necessarily guarantee your chance at the top of the private leaderboard. In the past, we've seen competitors end up overfitting on public leadership, and their model performance on the private leaderboard didn't turn out to be as expected. So be cautious, be patient and back your process/fundamentals!

Announcement:
 You should use
ONLY
the non-experimental features in the final submissions, but we’d be interested to see how including the experimental ones might impact performance. The experimental input variables in the dataset include: 'region__peak_sales_dollar_amt_per_hour_v2' and 'region__peak_returns_dollar_amt_per_hour_v2'. Please be cautious while building models and making submissions, as competitors adhering to the guidelines will only be considered for final prizes.

 

 

About The Challenge

 

Welcome to the Juniper Networks Global AI Challenge - an initiative by Juniper Networks for the community. We invite you to apply your machine learning skills and engage in a friendly competition with fellow participants of this AI challenge. 

 

About Juniper Networks

 

Headquartered in Silicon Valley, Juniper Networks is a leading enterprise in AI Networking, Cloud, and Connected Security Solutions. Juniper's AI solutions aim to solve complex problems in networking and/or security to the benefit of society. The Juniper Mist AI platform uses a combination of artificial intelligence, machine learning, and data science techniques to optimize user experiences and simplify operations across the wireless access, wired access, and SD-WAN domains. Alongside this, Mist AI offers next-generation customer support, which is the foundational element behind Marvis, the industry's first AI-driven Virtual Network Assistant.

 

 

Problem Statement:


A Retail Superstore chain wants to gain insights into the performance of each of its stores and make strategic decisions on its operating model. As a Data Scientist, you are required to understand the data provided and help the client predict the sales at the register in the current hour based on the sales attribute associated with each store.

The given data consists of 4 broad attribute types:

  • Register
    : Attributes related to the register at which the sales transaction has taken place

  • Cashier
    : Attributes related to cashier working at the register

  • Store
    : Attributes related to a particular store

  • Region
    : Attributes related to a group of stores in the region

Problem Statement:

Predict the sales at the register in the current hour and prioritize the individual attributes that affect the sales.

 

Prizes & Other Guidelines:

  • 1st Prize:
    $2000

  • 2nd Prize:
    $1250

  • 3rd Prize:
    $750

  • For positions 4th-10th: 
    $100 each.

  • All the participants

    must submit their notebooks/code, 
    and the final leaderboard will be released after looking into the code.

  • Note 1:
     The performance of the baseline model is ~80% on the public leaderboard data and ~37% on the private leaderboard data. Competitors' submissions will ONLY be considered eligible for prizes if they perform better than the baseline model on both leaderboards.

  • Note 2:
    Two input variables in the dataset are experimental: region__peak_sales_dollar_amt_per_hour_v2 and region__peak_returns_dollar_amt_per_hour_v2. You should use
    ONLY
    the non-experimental features in the final submissions, but we’d be interested to see how including the experimental ones might impact performance. This is optional, but you can include the results from using the experimental features in your notebook. To summarize, final submissions with experimental feature will not be considered for prizes.

Please note
that DPhi is a Belgian entity, and its competitions are open to all, except for the residents of states and countries sanctioned as per Belgian law. In addition, the intellectual property of the solutions will be transferred to the organizing company in exchange for the cash prizes.

 

 

 

 

Many thanks to our community partners for spreading awareness about the AI Challenge. Please feel free to explore their exciting initiatives for the AI Community.

This website uses cookies to ensure you get the best experience. Learn more